4:18
thank you [Music]
4:29
good uh good afternoon everybody I almost said good morning that tells Welcome Remarks
4:34
you how long I've been here today um we want to welcome everybody to 4:40
our experience you demonstration event I'm sure for the teams present today this feels like a long time coming and
4:47
so we just want to again recognize our eight teams who have put in a lot of effort and work to make today possible
4:54
we also want to take a moment to thank our partners and our sponsors for today 5:02
so today would not be possible without the support of the bill Melinda Gates Foundation they really bought into the vision of
5:08
experience you and invested in this to make today happen and we also want to 5:13
thank education Design Lab who have been great Partners throughout this initiative and have made uh today a
5:19
great event so we'll walk through today's agenda
5:25
real quick we're going to do a quick kind of intro into what experience you is 5:30
and then I will pass the Baton to Colin from education Design Lab who will kind of do a little 5:37
more of an overview of today's flow events and explain a little more in detail about who are the personas that
5:43
we're trying to reach with experience U H we've broken up the presentations into 5:49
four different use cases so we have a couple different use case groups that will be coming up and sharing their
5:54
demonstrations of leveraging artificial intelligence to create structured data for the workforce 6:01
after our first three use case groups go we'll have a networking break
6:06
we'll then have our last use case group come up and share their two presentations we'll then have a time for our panel of
6:14
experience you advisors to come out and share some Reflections and engage with the audience and then we'll have a little bit of a brainstorming session on
6:21
what can we do as next steps what needs to be done to take experience to you to the next level and what can we try and
6:28
Achieve together we'll then have some closing remarks and then we'll break and have a reception
6:34
and during today's reception uh there will be an opportunity to engage with each of the demonstration
6:39
teams so all the teams demonstrating today will have a table that you can go and gain a little more information in
6:45
detail about their projects and what they are doing and what their organizations are all about 6:51
we felt that 15 minutes for each of the groups was not quite enough for you to really capture and grasp the breadth of
6:56
what they're doing and we wanted to give them an opportunity to really explain and share with the audience what their
7:01
companies and organizations are doing around artificial intelligence and structured data 7:08
now some of you may have never heard of the T3 Innovation Network before I think a handful of the experience you
7:16
demonstration teams who are here today maybe hadn't had a lot of interaction with the T3 Innovation Network prior to
7:21
experience you so I thought I'd take this moment to just do a quick overview of what ex what 7:27
the T3 Innovation Network is and why experience you as an as an initiative that we are very interested in invested
7:34
in so the T3 Innovation Network has been around for about five years now 7:39
and the mission of the T3 Innovation Network is really to bring about the digital transformation of the talent Marketplace
7:46
and to do this we engage diverse groups of stakeholders including all of you who are here today 7:54
but the goal of the T3 Innovation Network is to really leverage this digital transformation of talent 7:59
Marketplace to make sure that all learning counts that skills and competencies can function like currency
8:05
and that Learners and workers are empowered with their data that they can explore new educational and Workforce
8:11
opportunities now the T3 Innovation Network is structured through four sub networks or 8:18
what we sometimes call the network of Networks we have our data and Technology standards Network our open competencies
8:24
Network our learning and employment records Network and then our jobs and Workforce Data Network
8:30
it's the experience you initiative is a chartered project through the learning and employment records Network or lern
8:36
as we call it within T3 and we explore projects and initiatives
8:42
that are really use case driven and standards-based and at T3 we try to be as vendor neutral as possible which is
8:47
why we brought so many different diverse groups of stakeholders here today we have a diverse set of uh demonstration
8:53
teams here today to really demonstrate that T3 is really about bringing different bodies together for the
8:59
collaborative good of this digital transformation of talent Marketplace
9:04
now if you're new to T3 and you haven't heard much about us or about the initiative we have a website that we
9:10
call the T3 hub and if you want to sign up for the T3 Innovation Network to join any of those 9:16
network of networks that might be intriguing to you or if you want to explore the vast array of resources that
9:22
we have curated on the on our website please visit the T3 hub or if any of your organizations have put
9:28
together resources and curated reports please let us know we're happy to publish them and post them onto our
9:34
website now
9:39
one of as we talked about where the T3 is a a use case driven initiative or and 9:47
Network and one of the areas of innovation that we've been at the Forefront of for a 9:54
number of years now is this concept of learning and employment records and how can we leverage learning and
10:01
employment records to promote the use case of skills-based hiring and advancement in the workforce
10:07
and so when we think about learning unemployment records we need to think about the next layer or next step of the
10:14
current types of employment records and education records that we have because if we just create another set of
10:21
transcripts or resumes that don't add any value to the system then the work at the T3 and Beyond is not going to add a
10:27
lot of value to the ecosystem and so how can we make sure that these learning and employment records or lers
10:32
are content Rich that they have that these are digital records that are machine actionable machine readable that
10:39
they have linked data that can help employers and education training providers identify and acknowledge the
10:46
types of skills and companies that an individual has and how can employers signal what they're looking for
10:52
so that's where we the T3 Innovation Network comes in and all of you come in it's helping us to find ways to embed
10:59
that richness of data in these learning and employment records to really facilitate and the advancement and Adoption of skills-based hiring and
11:05
advancement so
11:11
one of the things about experience U H and one of the things we try to do at the T3 Innovation Network is solve for
11:18
adoption gaps Within These use cases that we are exploring and so how can we solve for the ler
11:23
adoption Gap or what are the adoption gaps that exist so one of the issues with learning and 11:29
employment records or really ours is most of the digital records that are being issued today are 11:34
to current students or to somebody who has recently completed some sort of online or on-campus program or online
11:42
training program but there's tens of millions of individuals in the workforce that may 11:48
never get another credential in their lifetime so how can we meet the workforce where it is today 11:54
how can we create you know Mass amounts of learning employment records for the 11:59
current Workforce that can really help facilitate the adoption of learning unemployment records because if we wait
12:05
for there to be enough learning unemployment records for everybody in the workforce to really have that skills
12:10
retiring and advancement uh use case take effects and be it and
12:15
take hold is going to take decades so we have been exploring how can we get to 12:20
the point in the near future where everybody in the workforce can have these digital and machine actionable
12:27
records and so that's where we wanted to come up with experience you
12:32
we wanted to find a ways to increase the number of lers and also to find ways to solve for the 12:37
gaps that exist in the issuance of lers and then how can we get Lars to scale 12:44
now to do all this this is kind of the what this is the the what behind experience you're like what does experience you what are we about it's
12:51
about scaling and addressing the gaps that exist in the issuance of lers
12:56
but one of the most important things is is the how so how can we do this
13:01
and a lot of teams who are here today we have eight teams they'll show you different ways or different uh elements
13:07
of that how how can we solve for some of these issuance gaps and scaling gaps that exist 13:14
so what we sought to do when we started when we launched experiencing back in January is we announced that we were
13:19
going to explore the use of artificial intelligence or AI an Ali related tools like large language 13:26
models to take unstructured data about an individual this could be a PDF a website 13:34
a LinkedIn profile could be a resume and take that unstructured data
13:41
about that individual and convert that data into structured data that can be machine actionable 13:47
machine readable and then issued in the form of a learning and employment record and when we started this initiative
13:54
back in January when we launched we have been working on getting ready to launch for about six months and that was before
14:00
chat GPT kind of changed the world and made made this possible when we first started this back in almost a year ago
14:08
now I think when Jason and Naomi and Rachel and Kimberly started having conversations about getting this
14:14
launched we thought we might be lucky to have you know two three or four organizations 14:19
that would be able and willing to do this and so to almost triple those numbers to have eight teams today and we
14:25
had another 10 or so teams that expressed interest and were you know when capacity to do this we're thrilled
14:30
with the enthusiasm that the that the our partners across the ecosystem have shown and experience you
14:38
and so we're going to see today how these they're different applications of taking artificial intelligence to use
14:43
these unstructured data formats and convert them into structured data formats
14:49
and one of the things that we focused on um at the U.S Chamber Commerce foundation and T3 is you know the what
14:57
you know informs but the why really transforms like why are we here what is the impact that 15:03
we're trying to generate and Achieve through experience you and so if we can empower the workforce
15:09
if we can find a way to have incumbent workers unemployed workers military veterans be able to take their
15:15
experiences and get them these forms of structured data and learning and employment records as skills-based
15:20
hiring advancement begins to get more and more adopted across the across the workforce 15:25
there's going to be more and more potential for these individuals to become bigger players in the workforce
15:30
to have more employment opportunities to have more Education and Training opportunities and that why is what
15:36
drives us that's what that why I think is what drove eight of these teams to devote a lot of time and resources to be
15:42
here today just for the greater good a lot of these the teams that you'll
15:48
hear today from today eight months ago six months ago they weren't ready to do a demonstration on experience you but
15:53
they put in the time and energy and resources to do it and we're really appreciative of the effort that all of you have done to make today possible
16:01
so what we'll do right now is I'll I'm going to turn the time over to Colin Reynolds from education Design Lab he's
16:06
a senior education designer with education Design Lab he's going to come out and talk a little bit more about the
16:12
the flow for the different demonstrations and uh a little bit of more of a background on the last few
16:17
months around experience you
16:23
[Applause]
Experience You Update and Agenda Review
16:39
thank you well hi everybody it's so nice to see you in real life and not in tiny boxes 16:46
on our computer screens and when we're on those screens it's really easy to forget how many people are actually in
16:52
the room because they're just pages of them so it's nice to see you all here very excited to be on the stage and
16:58
sharing with so many brilliant people who have spent a ton of time and effort building some really interesting
17:05
solutions to many of the problems and things that Taylor just shared up here so we're going to walk through a little
17:10
recap of what got us to this point and then hand it over to our project leads 17:16
and their presenters to share some of what they've created over the last few months 17:22
um so I'm a senior Ed designer in digital transformation with education Design Lab education Design Lab works on
17:28
developing the future of educational work and we work in a lot of different areas around a lot of different things
17:35
it's hard to really like categorize exactly what Ed Design Lab does but in
17:41
in my time about a year working with the lab we've worked with a lot of K-12 higher ed employment groups most
17:48
recently HR Tech vendors to really try to understand and this is through some of some of my work is to understand what
17:54
are some of the challenges and obstacles related to digital transformation anything from learning and employment records to Tech Systems to people
18:02
systems to all of the parts that make data flow the Ed Design Lab also works 18:09
with higher education groups to create micro Pathways that identify alternative routes to using even alternative routes
18:16
are is shows you where we're at in terms of traditional and having something that's other those micropathways are
18:23
becoming the route for a lot of people to acquire skills and gain experiences to ultimately find a job or an
18:30
opportunity that that best fits them so education Design Lab we work with organizations all over mostly North
18:37
America increasingly more conversations with groups outside of of the states especially as it relates to Common
18:43
practices Global standards open resources and open open source tools
18:49
and and our our team really is it's Dynamic and diverse and one of the
18:54
things I love about working with this group is that no two conversations are alike and as Naomi Boyer likes to say we
19:01
solve Wicked problems and it's never easy every day we wake up and get to 19:06
work with people like you who have passion and and bright ideas and
19:12
opportunities for us to engage in collaborative work together so we'll dive into some of what we're here
19:19
for today experience you it really was meant to or sort of incubated to
19:24
generate lers at scale and Taylor mentioned chat GPT the timing couldn't have been better or more topical for us
19:31
to really engage wholeheartedly into some of this work and to help Elevate some of the value and meaning behind
19:38
these tools that's beyond just having a nice fun chat conversation where you can get something you know quirky and and
19:44
unique to send to your friends in a group chat which if you haven't done it yet it's hilarious your friends will love it and then you just point to
19:49
chatgpt but we're using Ai and these teams have developed solutions that are are much more impactful and are really
19:55
trying to transform the way that data is uh flowing and also Empower individuals
20:02
with data to seek value and find opportunity um
20:07
the promise of experience you from from the start of this was to create
20:12
Equitable access to lers for the masses education Design Lab focuses on creating 20:18
solutions that empower the new majority Workforce and and that is sort of loosely described as the the emerging
20:25
demographic of of humans that are taking on Workforce opportunities and are seeking jobs that are that are
20:31
continuing education and those paths aren't linear those paths as you all can imagine are as dynamic as the humans
20:38
that are experiencing them and so we're really trying to create a way to capture those experiences in a digital form that
20:45
is empowering and not just something that you know people type up in a word doc and put together in a resume and
20:51
hope that you know when they submit it they said the right words to to show up in the in the most common results so
20:58
that's what we're that's what we're here to do today we really want to empower individuals with their experiences in a
21:04
way that we know will or we hope will help create new mobility and opportunity 21:10
so how do we do that we really need to focus on the humans we created 21:16
um sort of three specific use cases that we think captured most of the population 21:21
of workers first being an incumbent worker somebody who is employed has 21:26
skills I I can identify with this group very clearly I was a teacher and administrator for 15 years I have a lot
21:33
of different skills but most groups outside of the K-12 world didn't think those translated into this job or that
21:39
role but I knew that they did project management was something that I didn't think about and then when I was put in a
21:45
context of being in a classroom with eight different group projects going on that I have to constantly be a part of
21:50
and very quickly realized I have a lot of skills to do something like that in another capacity so that was one of the
21:58
the use cases that we identified the second being an unemployed but experienced worker somebody who has a
22:04
lot of experience and this I think what came up a couple times were were people who have maybe a decade or more
22:10
experience in a specific area were laid off for whatever reason maybe a pandemic 22:15
maybe something similar an unexpected illness or life event and getting back 22:21
into the workforce was was difficult for them so that was another use case we identified and then our third for this was a military veteran somebody
22:27
transitioning into civilian roles and really trying to find ways to translate those skills that they've acquired
22:34
through their their years of service into something that seems employable 22:39
Seems attractive and that those individuals know actually can provide
22:45
value for an organization uh this fourth use case we we list it as optional and it ended up becoming a
22:52
pretty awesome project that you'll hear more about here today this idea of an Alumni network everybody in this room
22:57
has experience I'm guessing most of you don't have a digital wallet that has a bunch of verified credentials inside of
23:03
it that you plug into a talent Marketplace but wouldn't that be nice there are there are schools and and
23:10
universities employers now who are starting to package and create digital credentials new issuing badges over a
23:15
million badges have been issued so what happens with them where do they go those 23:20
are those are two separate things is taking my past experience it turning it into a digital record versus getting
23:27
some some sort of digital form from my institution that I'm participating in or 23:32
learning from today so that Alumni network use case became a pretty uh exciting and and scalable opportunity
23:40
for a team to work on uh a brief reflection of the timeline this as Taylor mentioned was incubated 23:47
just about a year ago we announced it in December we started working on sort of a white paper or draft paper to explain
23:53
what this concept would be and we put out uh we put it out into the the ethos uh in in January and had a big kickoff
24:00
event where we we've created an open form and said if this seems like something you might be interested in
24:06
here's a proof of concept or a list of technical requirements that we're starting to 24:11
um to sort of develop against or with and thanks to the Bill and Melinda Gates Foundation for providing sort of the the
24:18
funds to to seed this conversation to seed these opportunities and it was an 24:23
open call and groups that were interested um and there wasn't a promise of any sort of cash reward there wasn't any
24:29
development dollars put into this so these are groups that were very much interested in this work and wanted to
24:34
invest their their time and resources to make it happen um and and just in the spectrum of 24:39
groups that you'll see today we've got teams that were bootstrapped as like Pierce startups for this and then we
24:44
have some very well established mature and recognizable organizations so really awesome to be at the center of this and
24:51
and really great to see that those two sort of ends of the spectrum are brought together for this project and the last
24:57
six months has been weekly meetings those of you that have been on them uh you know that I had plenty of quirky you
25:04
know random jokes and water cooler talk but we waited for a couple more minutes for people to enter the room before we
25:09
hit record and actually talked about the project work that we were doing and so to be up here today to give the stage to
25:16
those project leads and and developers and presenters to talk about their work is really an honor after this there will
25:24
be a recap of all this that is put together in a nice formal paper that will be published and shared with this
25:30
community and the broader skills community foreign the process was messy this diagram makes
25:36
it look nice with you know square corners and 90 degree angles it's not that it just looks nice on the screen
25:44
the projects really are diverse they are challenging and the collaboration across 25:50
organizations across groups was quite impressive that is the kind of work that excites me the most is being in uh in a
25:58
room in a conversation with people who represent different organizations different populations but they're
26:04
working together to try to solve problems and this problem was mass unstructured data that we want to
26:09
structure we want to use Ai and other emerging Technologies to do that now is a great time to do that if you're trying
26:15
to use AI 10 years ago it might have been a little more difficult now Chachi PT has you know the the ease and
26:22
function to plug into a platform that you're developing so we're excited to 26:27
see what the teams share here today there's really kind of three uh forms of 26:32
AI that we put out into this conference station to begin with and that more or less there are different flavors of this
26:38
that you that you'll see and our leaders will will share with you the first is credential mapping 26:45
um the taxonomy conversation the skills Library conversation uh is tricky when you talk about sort of um matching what
26:52
your skills are with this sort of you know repository of skills and definitions what what I Define as a
26:59
skill what you define as a skill might be different so how do we really start to map some of those things against each other extracting skills from a resume
27:07
from a cover letter from somebody's experience from somebody's spoken word from a demonstration of their experience
27:12
maybe a a unofficial transcript or any other evidence of learning AI is really 27:19
powerful and helpful in doing that and and the third being this idea of an AI agent so this is much more of the
27:25
conversational type chat bot experience where you're sort of having a conversation you remember clippy who
27:33
used to sit in the Microsoft Word doc could kind of help you along your journey so this is that sort of version of it some some way or some agent being
27:40
able to engage the individual to say hey you said this which kind of sounds like that which Maps towards that is that the
27:47
skill you have you can say yes or no but it's a little more Dynamic a little more conversational in nature and there's a
27:52
really awesome example that will be shared I put these up just because this is a 27:58
technical demonstration and I want to be uh I want to offer the reminder that these were the the check boxes that
28:05
needed to be hit in order for teams to be up here today although you may not see every single piece of this in every
28:11
demonstration know that these were behind the scenes in the conversations and ultimately what each team was
28:17
required to hit to be up here today the the structured Json LD formats is a is
28:24
still a very emergent and um solution or or structure for these
28:29
for these data sets some some groups very proficient in it others this was brand new to them so those those three
28:36
schemas that are they're up there one at Tech is is one standards organization 28:42
who has been putting out uh verifiable credential compatible standards that that groups were using so they were
28:48
definitely helping out with some of the conformance testing in the background 28:54
we also had other groups step up and support this initiative in different ways by offering their uh expertise and
29:00
their time and to run some workshops with us and these were part of our weekly or bi-weekly meetings IBM led a
29:07
couple of workshops around ethical hacking and this idea of the impact explicit or unintended impact of some of
29:15
these Technologies on the humans so as uh we're walking through the use cases today there will be some personas or
29:22
example personas that were developed by IBM in collaboration with one of these teams dexterra Institute offered a
29:29
workshop on data management those of you that know dexterra they this is their area of of expertise and they really
29:35
work on Integrations with groups and across systems credential engine was always in the 29:41
meetings Dev and pretty sure she's here somewhere she was always ready to have conversations and support project teams
29:47
with any questions they had related to credential engine and their credential 29:52
transparency description language which we always just say ctdl so for me to say that out loud was a bit of a challenge
29:58
and then yeah it's being blocked on my monitor here but Acro was supporting 30:04
um uh a that fourth use Case by creating a cohort of registrars who are having
30:09
conversations around what they would find valuable to be extracted from transcripts and credentialed for
30:16
individuals who could be transferring could be graduating who might have partial degrees or partial credits kind
30:21
of hanging out there so those groups supported this kind of in the background and they won't be presenting today but want to acknowledge their time and
30:27
effort so for today before I introduce our first team the flow is as Taylor kind of 30:33
outlined will be built around this idea of a Persona this idea of a use case the 30:38
disclaimer at the bottom there these are just examples this this doesn't mean that we're identifying or that every
30:45
person identifies very explicitly or clearly with one of these personas there's they're derived from 30:51
conversations with humans who personify this this sort of concept of the
30:57
experience and learning so that's what will be driving our use cases today 31:02
and the flow of of the project teams uh one thing I should mention uh two things to mention 31:08
um throughout the uh the conversation today on your table there are two QR codes one QR code if you scan will lead
31:15
you to a conversation on LinkedIn a sort of back Channel that'll be happening throughout the day so uh instead of
31:21
using some social media platform and pitting you know two big wigs against each other we just went a different
31:27
route and we're gonna have somebody who's facilitated a conversation on a different social media platform but that
31:33
was sort of the we thought the the lowest common denominator for today so if you want to engage in that
31:38
conversation online those of you that are watching online please head over to LinkedIn and look for the micro
31:44
credential microverse and there will be a conversation going on the other QR code is a simple feedback form and if
31:52
you have any feedback that you would like to offer any of the teams that are up here today including the education
31:58
Design Lab or the chamber the T3 Network that is the form that you can fill out as many times as you would like you'll
32:03
have to select which team you're giving that feedback to but feel free to engage with us and provide feedback that way
32:09
there also is a spot in there for you to ask any questions or offer any questions that you would like our our panels
32:15
panelists to reflect on or respond to during our group conversation
32:22
all right I think I got it all I'm sure I'm forgetting something
32:29
our first use case uh that's going to guide our conversation today is around this idea of the incumbent worker
32:36
somebody who has skills who uh is looking maybe for a different job the 32:42
Uber driver is an example that somebody who may have be new to that profession 32:48
because they're sort of biting their time while they look for something that they know they're more qualified for so part of the personas that we think is
32:55
really important and that the IBM team helped to sort of flesh out where the what this person is feeling what they're
33:02
thinking and what they do I think those really putting yourselves into those individual shoes to to conceptualize
33:10
what their experience might be as they are trying to get their experience into a credential format what does that look
33:17
like what is it what does it feel like what's this individual thinking so that'll set the table for our first use
33:23
case around the incumbent worker a lot of these Solutions cover more than one use case I know we have a team up
33:30
here that's going to talk about some of the the testing services they provide that can transcend any single use case
33:36
so again these were meant to provide these teams opportunities to talk to and 33:41
interview individuals one-on-one to better inform their development process 33:46
so with that said I'll introduce our first group our first group is go Beckley gobekwe is an early stage
33:54
pre-product startup building a universal Talent passport to put the experience of individuals first in the emerging ler
34:00
ecosystem by enabling people to create for themselves the talent data organizations want to adopt Gobekli
34:07
believes that if everyone had an easy valuable and rewarding experience to help them collect contextualize and leverage their entire work and education
34:14
record they could be better equipped to understand represent and grow themselves through each stage and organization in
34:21
their life Journeys the Project Lead for this is Danny Doan he is the founder and CEO of Gobekli he's a homeschooler
34:28
turned National College leader he studied comparative history of ideas he founded built and sold a top-ranking
34:34
Google marketing agency before founding gobekwe to solve Talent visibility challenges for everyone presenting with
34:41
him will be coyote ezeke he is also a co-founder and CTO of Gobekli he's the 34:46
Chief Architect and the developer of solid verifiable credentials prototype under Tim berners-lee at MIT
34:52
he's also an expert in decentralized identity and ler Technologies so with 34:57
that said we'll kick it off with our first team thank you
35:04
[Music]
Use Case Group 1: Incumbent Worker
35:16
thank you hello everyone my name is Ike and I'm
35:23
thrilled to be here with Danny in this first of his con event here in a nation's capital my first time 35:29
personally and we're excited to be here to discuss a very important topic and that is the role of AI and
35:35
representation and opportunity creation for students and job Seekers
35:42
hi everybody I'm Danny Doan very happy to be here we just want to say thank you to the U.S Chamber of Commerce
35:48
Foundation the bill and Melanie Gates Foundation everybody who put this together we're not going to waste too much of your time we've put together
35:53
actually a 14 and a half minute video for you guys so we're going to let it do the talking and uh and then at the end
35:59
of it we will be around for questions and have a table over there at the end so without much further Ado I'm just
36:06
gonna by the way there's US Special thanks to the University of Phoenix who helped us with our use case and testing
36:12
and you'll see some of that here in a second quebeckley's founded on the idea that 36:19
everyone should have an easy way to collect and control their entire work and education record in an app that
36:25
helps them understand represent and build themselves as well as partner with organizations throughout their entire
36:30
lives we think what's needed isn't just a credential wallet but something we call a universal Talent passport a class of
36:37
app that we believe will catalyze global ler adoption by putting the experience you the individual has above everything
36:44
else I have some neat new skills and some classes I took and I'm not 100 sure how 36:51
I want to incorporate that into you know things like resumes and where where do I talk about it and where don't I myself
36:58
for example I've I've picked up who knows how many different skills from for the last 20 years being in the military
37:04
yeah I just finding a way to incorporate that into my resume but like I'm also 37:09
looking forward to the future where you can put in your resume and then throw 37:15
your badges up there with it to provide further Evidence when we were approached about joining
37:21
the experience you project we saw it as an opportunity to test the app with align Partners interested in pushing the
37:28
ler ecosystem forward for everyone we'd like to thank the University of Phoenix for their support and their
37:34
collaboration on this project their technical logistical and human support on this have been essential to
37:40
making it happen [Music] we'd also like to thank councilor Kendra Laura from the city of Boston and BJ
37:47
usagu director of Civic Civic engagement for their help collecting additional 37:52
feedback from non-controllable users through the Boston University of Phoenix provides 37:58
opportunities for working adults to access affordable career relevant education as they pursue their careers
38:05
we're constantly looking at Trends and new ideas to understand how best to serve our students and graduates in the
38:11
years ahead one idea that seems to be convergent is that in the near future students can carry all of their own
38:17
records in a comprehensive digital format in a decentralized way so Ingo Beckley approached us with their
38:24
design for a passport that used lar technology as a super wallet the idea matched a lot of what our team had been
38:30
hoping to see [Music] focus group gave us an opportunity to talk with real people from diverse 38:36
backgrounds and getting their feedback on the concept of the ler ecosystem and wallets and pass Sports in general
38:42
before we even showed them our design we view the emerging ler ecosystem at 38:49
the global Highway for the growth and mobility of talent in this sense the talent passport is a 38:54
vehicle for individuals to navigate toward their own ideal futures
38:59
in the focus group we first present to the ecosystem and its goals as well as a 39:04
concept of passports and wallets and their intended purpose these are some of their ideas and 39:11
feedback if you guys have the Magic Bullet that will put an end to the upload your 39:18
resume and then retype your resume into all of these discrete Little Boxes I'll 39:24
sign my 401k over to y'all like if you've got all this information about about my skill set I want to know
39:32
if I'm a good fit for this job well how about some something to our uh it cross 39:39
talks between different apps like crosstalks between LinkedIn or
39:46
um yeah maybe even the the school's website where you've got skills already listed 39:51
if it populates here let it populate there yeah
39:57
so I think one of the things I would find valuable in this type of technology I have probably no less than 45 versions
40:06
of my resume right because you adapt it um based on what you're applying for and 40:12
the skills needed for that job and so um there's not a way just to uh check 40:19
boxes and say these are the skills I'd like to include or these are the skills that like omit like these skills don't matter just you can leave these ones off
40:25
for this specific job and in doing so I fear that I lose things like I lose data 40:31
points from 15 or 20 years ago because I removed it from a resume and maybe didn't save that copy and just don't
40:38
think to re-add that from a job I had you know when I was 24. so so I think 40:43
that having a way to have that data ever present and then select what to put 40:48
forward without losing the Integrity of what you're doing would be helpful 40:53
based on these and many other related needs we designed an easy and scalable experience to allow anyone to build a
41:01
complete Talent profile that's beneficial for themselves as well as their organizations 41:06
our Theory going into this was that this requires three basic components 41:12
first passport pages to store all evidence and manage filtered profiles
41:18
and data sharing permissions with organizations second a personal talent tree 41:24
to easily organize and visualize all aspects of your talent data and last but not least a conversational
41:31
companion to help you collect and expand your data to better understand represent 41:37
and grow yourself we so experience you as a chance for us
41:43
to build test and receive feedback on our theory on a national stage with some 41:48
great partners we see the talent passports as a way to help students collect contextualize and 41:54
fully Leverage The comprehensive learner record and the badges we spent the past 24 months skills mapping and issuing
42:02
we saw gobekli's project for experience you as an opportunity to push ourselves to the next level of innovative student
42:09
services as well as a chance to support entrepreneurship and advancement in the lar ecosystem as a whole
42:17
we've recently completed implementing badges as well as our own version of the one edtech CLR so it was exciting to
42:24
meet a team building a passport that can actually adopt and use this data so we can extend its usefulness to our
42:29
students we spent the past two months working with Beckley to learn and test the VC 42:35
API chappie protocols to allow a subset of users to be able to collect the badges in CLR we are already issuing for
42:43
them for University of Phoenix students The Experience begins with downloading their 42:48
skill data in the form of open badges 3 from their student dashboard to the passport 42:54
thereby satisfying experience you criteria 1 and 7 at the same time
43:00
incidentally University of Phoenix's pre-existing work creating and packaging skilled data also satisfies Criterion
43:07
two our app is Guided by a conversational companion that helps you to accomplish 43:13
your goals in two to five minute dialogues for our project we set up four of these 43:18
conversations to guide you through the purpose and scope of this demo and provide you with our core Talent
43:25
services in the future we will add a queue of AI controlled conversations based on your 43:31
needs and goals over the past year and a half we've been
43:36
collaborating with blind spot AI out of Prague to develop the next generation of skill tagging that we call stamp
43:43
which normalizes Talent data against cross-walked ontologies
43:49
the conversation experience prompts the user to upload the PDF of their resume 43:54
we worked with coyote to help build a method using openai to interface with 44:00
allnet and return if set of Talon categories expected to be associated
44:05
with a given role or experience described in this prototype we are returning 44:11
knowledge skills and activities categories for the skill tree and only for resumes or exported LinkedIn
44:18
profiles however using the same method and algorithm we can use the conversational 44:24
AI to collect and gather enough context for us to understand experience and 44:30
extract Talent from any kind of artifact story or credential the user provides 44:37
they believe that what we have demonstrated with measuring skills from a resume satisfies criteria 8 skills
44:44
data enhancement talented deduced by AI from odet terms
44:49
is returned to the user where they are asked to scale measure and normalize the Knowledge and Skills and activities
44:55
based on their personal experiences this method of quantifying the skills from the past experiences is explained
45:01
by the conversational agent in the app as the users invited to help to help the Abacus tool satisfying
45:08
requirement for also the fact that we've chosen to have a conversational based design in the 45:13
first place was out of research on how best to create the most human-centered design possible with the simplest most
45:20
accessible scalable and adaptable interface to allow for different people's goals and needs this also is
45:26
explained to the user in the beginning conversation satisfying requirement 3. the output data is presented in neat
45:33
visualizations that reside in your personal talent tree additionally we provide a utility to 45:39
convert this Talent data into open badges 3 credential on demand satisfying Criterion 2 which calls on us to convert
45:45
under structured Talent data into admissible Json LD format between ourselves and University of
45:51
Phoenix we are pleased to report that we have fulfilled all of the technical requirements of the experiencing project
46:00
in addition to accomplishing the technical requirements of the experience view project with both credentialed and
46:05
non-credentialed users our goal was to measure the user feedback of the three key pillars of our design after seeing
46:12
the design of our AI companion we're calling picia the conversation queue and how dialogue can help them accomplish
46:19
micro Journey goals of various types focus group members rated their comfort 46:24
using AI in this context a minimum 700 everyone in the focus group said they would use the app at least on a monthly
46:31
basis with over 55 saying that they actually would use on for weekly things like discovering education opportunities
46:37
personal growth skill enhancement Career Development work-life integration mentoring coaching Etc
46:44
we also found that the perceived value and usefulness of the talent data visualizations organized into a personal
46:51
it's very promising the vast majority of people rank Talent data visualizations a 10 out of 10 of usefulness for
46:57
self-understanding and representation the same was true for clarinet of the organizing the talent into a tree-shaped
47:04
mosaic of interconnected Talent data categories when they asked what they would like use this data for they said
47:10
that it would be for career exploration decision making personal branding performance School tracking networking
47:16
self-motivation and confidence goal alignment career progression and more 47:22
finally the third and final essential piece of buy-in we were looking to measure was how people felt of the
47:28
passport Pages for organizations while it isn't included in the Prototype we showed our plan passport Pages for
47:35
cataloging evidence through stories facts and credentials and how they'd make and manage their variable use of
47:41
profiles contextualize the third elements the passport Pages for organizations that 47:46
mediate the way the two-way exchange of talented the response from participants indicated 47:51
over 90 Comfort was sharing their data with chosen organizations in exchange for job search assistance resume
47:58
building professional development career coaching internal growth opportunity alumni programs and more so long as they
48:05
had control over which Services their data was being used for experience you provided us with a clear
48:12
use case to focus on the line Partners a deadline and a directive that we needed to put our 48:17
research and designs and new technologies we've been working on over the past three years into the
48:23
user-facing Prototype that we could finally test this prototype has already given us the 48:29
validation we needed to commit to a full beta build by December to open up for six Alpha client Partners who want to
48:36
Pilot passport pages with their people and have a say in how this technology and service evolves for everyone
48:43
we've already spoken with key hris SAS and LMS providers who have confirmed they're ready and willing to build
48:49
Integrations to leverage passport page data should their school and employer clients wish to do something
48:55
those schools employers we've spoken with have told us they would be interested in doing Pilots to explore
49:01
what passport Pages can do for them to improve hiring management and growth of 49:06
their people if such a passport from our research we were convinced that 49:12
if people were an optional source of data for their organizations that was more comprehensive and more reliable
49:19
than anything that they could collect or verify on their own they would invest in systems that could integrate with and
49:25
leverage that data in Mass the emerging ler ecosystem has created
49:31
the framework to make that possible experience you the individual needs 49:38
first creating the data organization's need to become better partners with everyone
49:44
reach out to us if your organization is interested in becoming an alpha client piloting the passport page in our beta
49:50
launch this winter we are Delaware C corporation and currently entertaining funding 49:55
opportunities to accelerate our timelines so if you're interested please finally we would love to join or help
50:02
build any Coalition promoting the needs and rights of users experiences on the ler ecosystem if you're interested in
50:09
discussing ways to collaborate or partner in research Innovation ethics or standards for the experiences of users
50:15
in our emerging ecosystem we would love to connect thank you very much to the U.S Chamber 50:20
of Commerce Foundation the education Design Lab and the Gates Foundation for putting this together we are extremely
50:26
grateful for the opportunity to share with you all what we've been thinking about and working on for the past few
50:32
years and to be able to share with you all where we think it can go please stop by our table we would
50:39
[Music] love [Applause]
50:49
thank you everyone this last slide is just the validation for the Json LD
50:54
files just per requirements and uh yeah if you want to talk with us we'll be at 51:00
the table in the back and you can reach us between sessions but we're 30 seconds over so we appreciate your time thank
51:06
you [Applause] thank you
51:11
[Applause]
51:27
that that was awesome that was more like what Danny looked like every time I talked to him with the you know in his
51:32
office and the backgrounds and everything um it was a great example of sort of a comprehensive uh project working towards
51:40
all of those uh proof of concept requirements and a project that was also 51:46
kind of built around this project some of the others that you'll see today sort of uh integrate parts of these uh
51:53
requirements into a bigger solution but in the background creating some solutions and data structures that
52:00
provide for interoperability is the goal of all of all this next up we have a uni 52:06
uh uni is a website which offers its users the ability to explore classes find jobs and build custom career
52:12
Pathways currently on uni Educators and employers can post their class or job and have it automatically matched to a
52:19
growing database of Rich skill descriptors or rsds what makes uni different from other job boards and
52:25
class marketplaces is it's focus on connecting education and experiences to outcomes and encouraging the direct
52:31
connection between a network of employers and educators the Project Lead for this is Jake 52:36
debatista he is the CEO of uni he is an active mentor to students at the College 52:42
of Charleston and a member of the Charleston's startup and web 3 Community 52:47
he is passionate about emerging Technologies and their impact on the future of work in education and that's
52:53
stemming from his frustrations with the current inequities and cost in higher education 52:58
he has worked for multiple startups in the public good space and most recently at Walmart where he actively contributes
53:04
as a member of the Sam Sam's Club technology team as a ux researcher so Use Case Group 1: Incumbent Worker
53:10
coming up to the stage will be Jake dibatista with uni [Applause]
53:16
[Music]
53:32
hello everyone and thank you for the intro my name is Jake I am the CEO of uni and I'm going to talk to you about
53:40
what we've been building over the last six months as part of this competition 53:45
I'm going to start off by telling you a little bit about what uni is and why we started the company I'm going to go into
53:50
a little bit about what LMS do and then we'll go right into a product demo 53:57
we're going to wrap up with some things we've learned over this last six months about llms and the space in general and
54:03
then go into a little bit of a product timeline so opening today I kind of wanted to 54:09
talk about this quote that I saw recently and it's four for four not four
54:15
for forty and what does that mean well you know I think when colleges were first created there was this Paradigm
54:22
that you would go to school for four years you would learn how to be a manager of a plant or do some
54:29
profession and that would last you 40 years and unfortunately I think that we are 54:35
now in a time where this is no longer applicable I think that reskilling is going to be 54:40
more in demand as technology moves the markets move and we're going to need to 54:45
know how to reskill and re-employ ourselves constantly to stay competitive in this market 54:52
so what's uni yeah that's where a little Brash we're a startup
54:58
um Uni's dedicated to helping students take classes that will help them
55:04
in their career me and my co-founder actually met at John Hopkins University and to be honest we were pretty
55:10
frustrated with our experience a lot of the classes just didn't connect to the real world and we couldn't
55:17
understand why we were taking them and we had to know that there was others 55:22
like us out there that wanted to only take those classes that actually would teach them something that they found
55:28
valuable or more importantly the market found valuable so our idea behind uni is to have a 55:34
place where industry could meet education we did this using a repository of rsds 55:41
which are rich skill descriptors this is a really great way to document skills and use them and our idea is that on our
55:50
platform teachers could come to show their classes off and Market them and display them employers could come there
55:58
to recruit and connect with these teachers and students could actually plan their careers with education and
56:05
Industry in mind instead of having two isolated sources of data
56:11
um so let's talk a little bit about lom's you know about a year ago when me and my co-founder met LMS were not the
56:18
hot thing in fact blockchain was if anyone still remembers that but um it's still something a technology we
56:25
believe in but uh regardless wire llm's important to us and why have we chose to 56:31
kind of pivot and take part in this contest over this last half year well first of all let's look at a fact 56:36
over half of resumes submitted online never make it to a real human that was 56:42
off a report recently published from indeed that's because the what's on a resume isn't usually machine readable
56:48
and in fact they're kind of looking to check buzzwords or keywords which are self-verified and a lot of people don't
56:54
even know to use so that's a big problem second of all cooperation is a huge 57:01
problem right you have all of these employers that have their own definition of skills their own hiring needs and you
57:07
have all these colleges and they're not only competing amongst each other they're not really talking to one another it's it's kind of these isolated
57:13
ecosystems so as a student you have to figure out not only the best school for you but how that school connects to a
57:19
job that you actually can can place into and want and that's a problem when these are isolated because you get redundant
57:26
skills or skills that don't really map to careers
57:33
um so what did we do about it well we built this product uni and the first part of uni is a place for 57:39
teachers to post their class this is still a young product so if you see any design glitches or anything give us a
57:46
little bit of benefit of doubt it's getting better every day but what I'm showing you here first is just a simple uh class post and what you
57:53
can see here is at the very bottom we have rsds connected to every class that goes into our library
57:58
this can happen manually by a teacher going in and choosing an RSD from our library 58:04
or you can use AI to actually read your class syllabus or class posts and 58:09
recommend rsds for you that we think it's important that teachers understand what an RSD is because if they don't I
58:17
don't think there's a good way to understand what they're teaching towards so the first step here was to make a tool that we think that teachers will
58:23
love and use to Market their class and as well as show their class to their students 58:28
the next thing we did here was make a tool for employers I think this is the hardest part of this whole problem is how do we get employers to buy in so
58:36
what you're going to see here as an employer who can literally just write into an open text box what they are
58:41
looking for in their ideal employee they can write that they're hard working they can write that they've worked for 20 years in Tech and it will actually build
58:48
out in real time a job post with them making a call to open AI um so on here I've kind of just shown
58:54
you me making up a prompt in about three to five seconds it's going to go ahead and fill out this entire form for me
59:01
which is pretty cool now building job post is nothing new and
59:07
we've been doing that for years but what we think is really cool is the ability to actually tag this job post with rsds
59:13
so again you can go ahead and search our library for an RSD or you can go ahead and create your own
59:20
so what we're going to see here up next is that this job post didn't actually have an RSD that matched the job and
59:27
this employer now is going to type in what skill they're actually looking for now why is this important well
59:34
libraries over time become thirty thousand entries a million entries when everyone keeps entering the same skill
59:39
over and over again what we can do with AI is we can actually match your natural 59:45
language of what you think the skill is to other skills in the library or recommend a new one and create it for
59:51
you so that's what you're seeing here this job now has a skill tag to it which was created on the Fly using Ai and
59:59
again the whole goal of this is to save people time technology shouldn't become a burden it should actually become a
1:00:04
Time Saver help you make more money for your business help you hire better talent and we believe that this does
1:00:10
this by by mapping to these skills so let's kind of get to the meat and potatoes of our presentation which is
1:00:15
how does this actually work for students because as I said in the beginning our whole company was built to help students
1:00:22
so the last part of this now that you're going to see is I went ahead and I just made a fake resume in Bard but you could
1:00:28
really use any natural text Data you want about your wife about your career
1:00:33
and what we're going to see here is how uni can now read this resume and actually map it to the rsds used by all
1:00:41
those jobs and classes I just saw you getting created to make recommendations that if you want this job then take this
1:00:47
class essentially answering the question of why am I even taking this um so right now you're seeing that
1:00:54
resume getting built and how the student's going to go and
1:01:02
generate a resume with AI
1:01:12
that resume posted in there
1:01:17
and now they're going to be selected a recommended a list of rsds that are relevant for their resume so instead of
1:01:24
Now using a resume on our site they just use rsds that's it you don't have to do anything else you just select the ones
1:01:30
that work for you and as you now start taking classes and applying jobs this builds as a profile for you and what's
1:01:36
really cool is that every time you apply for a job and they actually hire you it verifies the claims of all of those rsds
1:01:43
you've selected so in that way you're now having this profile or this data point that this employer has vetted and
1:01:50
validated for you and we think that long term there's a lot of cool things you can do with this and this is just really the start of
1:01:56
where this our company can do so let's talk a little bit about what we
1:02:02
learned first and foremost we wanted to make an interface that we could give to anyone I wanted to be able to go up to
1:02:09
my friend or my cousin who's currently looking for a job right now and tell them hey I know you're frustrated I know
1:02:14
you're on indeed I know that you spend all this money on education and you're still not really finding the jobs you
1:02:20
want I have a tool for you that was the most important thing for me because if I can't hand you a tool that's you know
1:02:26
intuitive and easy to use I don't have any feedback loops on what to build next so that was the first thing that I
1:02:32
really set out to do was to build a real tool that a real person could use uh second we've learned a lot about Ai and
1:02:39
using the open AI function calls and AI calls that are out there right now first we learned that if you have visuals or
1:02:45
tables on your syllabus your resume it's still not really good at learning that right it's still really hard to take a
1:02:52
PDF that's really fancy and has a lot of dressing on it and unpack it into skills data 1:02:58
um second we learned a lot about this thing called function calling open AI is constantly releasing new tools to make
1:03:05
it easy to build products on their platform function calling came out about two months ago and it's dripping through
1:03:11
our product what function calling does is instead of just saying hey open AI give me a text-based answer it actually
1:03:17
will give you a functional answer like a Json object or actually let you call 1:03:22
an endpoint or a function that you've created in your software so you can do way more now with open AI than you could
1:03:29
say even four months ago in fact you can even now fine-tune open AI so that when 1:03:34
it searches an RSD it will specifically search your library of rsds not the entire internet or try and make up its
1:03:41
own hallucination of one it's incredible in the last six months what openai has been able to do and lastly we've learned a lot about
1:03:48
modeling data and how data redundancies need to be prevented and how to get data 1:03:54
clean as you get more employer more job and more student data in this thing we also learned a lot about storing data
1:04:00
with our friends at the learning economy and at best AI and how we can map this data over time and that's something we
1:04:06
want to do a lot more with as we now have a framework to build off of
1:04:11
so what is our timeline over the next six months here as we go into the back half of the Year well as I told you
1:04:17
we're a startup we literally didn't have unid.com until April of this year and until probably three weeks ago the whole
1:04:24
skills Library piece of this and generating skills wasn't even a thing so we're building really fast and we're
1:04:30
constantly iterating based on user feedback starting in September of 2023 we're going to have learn cards set up
1:04:36
so not only can you come in and copy paste text but if you have a digital wallet it will be able to actually
1:04:42
connect to our platform and you'll be able to extract all those skills out of there so you don't even have to do the resume or the text-based write-up piece
1:04:49
of this anymore and lastly we're already working with some recruiting platforms and thinking about how can we tie an
1:04:55
active job posts and feeds so that these rsds actually get weighted so we can say hey instead of waiting for Corsair to
1:05:02
publish a job report we can tell you there's 50 000 job openings these are the three rsds that are most commonly
1:05:08
used and here's the classes they're actually tied to those rsds and real time gives students a leaderboard of
1:05:13
what to enroll in and select rather than leaving them their own you know out in the wild on their own
1:05:20
so you know for the last three minutes here I just kind of want to close um with one last quote and it's the
1:05:26
whole purpose of education is to turn mirrors into windows it's kind of an interesting quote if you 1:05:31
think about it when you're looking at education it is really in some ways a mirror of 1:05:38
yourself right you choose to take these classes you're doing it for yourself I mean you get a grade and obviously a
1:05:44
diploma at the end of it but the whole idea of it is you walk Inward
1:05:50
and at the end of it there's a vision that you're working towards and I think that's how education 1:05:56
should work and that we've lost a little sight of that it's become so obsessed with the transcript or just the the
1:06:01
ecosystem itself that we forget that the goal of this is often a better life or a 1:06:06
wife you're a little happier or can do something that you love and we need to get back to that we need to move past
1:06:12
you know wherever Funk we're in now and really make it student first and that's 1:06:17
the goal of my company and I'm hoping that AI can make a difference in this space so thank you all for your
1:06:22
attention and give you two minutes back your time and uh hope you have a great rest of the week [Applause]
1:06:52
of a soft spot for the web3 community um because I've spent a lot of time working in it and collaborating with
1:06:59
people like like Jake and it's the a lot of the Technologies or some of the the 1:07:05
layers underneath are the same um but like you said blockchain sort of got a 1:07:10
whatever you whatever you think of when you hear the word blockchain that's what happened to blockchain so some of those
1:07:17
Technologies are still being developed on and built for uh but I want to highlight too just the collaborative
1:07:22
efforts that were put together and that first uh demonstration the University of Phoenix was called out in in the uni
1:07:29
demonstration you call that learning economy foundation and best AI part of that open call that we put out at the
1:07:35
beginning of this um this project was for anybody who was interested in supporting or developing
1:07:41
in any way we had a number of groups that weren't able to commit the time and or resources we we were also pretty
1:07:48
lucky to get a number of individuals who said something similar so there were other ways to support this initiative by
1:07:54
partnering with the group advising with the group we have a group of advisors that showed up to calls just about every
1:07:59
week made themselves available for one-on-ones or really any kind of resource support that they could provide
1:08:05
even if they didn't lead a project team that was in this demonstration so I want to be sure to acknowledge those
1:08:10
supporting organizations and the individuals that that represent them for their efforts um in what you see up on
1:08:16
the screens here today uh so we're going to transition a little bit into our next uh 1:08:22
uh use case uh use case number four we're not going in in numerical order which is the Alumni network so Case
1:08:30
Western Reserve University uh collaborated with the University of Pittsburgh and Chapman University and
1:08:36
with a little side of Columbia College in there to work on this next project and really 1:08:42
focus on taking transcript data and transcript information and skillifying
1:08:47
it uh so Case Western Reserve University is a private research University in 1:08:52
Cleveland Ohio the x-lab which is where this was incubated out of at Case 1:08:58
Western engages with industry Partners to develop knowledge and talent for responsible digital Innovations working
1:09:04
with a team of multi-disciplinary students faculty and other centers throughout the university x-lab develops
1:09:11
responsible technology Frameworks and tools and works with its company Partners to design new digital
1:09:17
Innovations for products services and business models a project lead for this 1:09:22
team is Young Jin Yu young Jin is the Elizabeth M and William C true haft 1:09:29
professor of Entrepreneurship and professor of Information Systems in the department of design and Innovation at
1:09:36
the Weatherhead School of Management at Case Western University he is also a faculty director of the
1:09:42
x-lab he's a WBS distinguished research environment professor at Warwick 1:09:48
Business School in the UK and an AIS fellow he has worked with a variety of leading global companies on digital
1:09:55
strategy digital transformation and Design he will be presenting with ishan Gupta 1:10:01
ishan is a recent graduate from Case Western Reserve University and he graduated with the BS in computer
1:10:07
science specializing in artificial intelligence he graduated in May of 2023
1:10:13
and has been a part of the x-lab at Case Western for over two years where he worked closely with blockchain personal
1:10:19
data stores and AI based solutions for helping Drive the digital Innovation efforts of the organization he has
1:10:25
worked at startups as well as big leading companies including visa and Highland software so this project is
1:10:31
from the team at Case Western a Reserve University [Applause]
1:10:39
[Music]
Use Case Group 4: Alumni
1:10:54
foreign well thank you for having us here it's exciting to be here to share our work my 1:11:02
job here is to uh to set the stage up and then introduce our star uh recent 1:11:09
graduate student who's going to join JPMorgan
1:11:14
next month and this is the last project he worked on with us so I'm very proud 1:11:21
of what he has been able to do so I'd like to also I'd like to recognize my 1:11:28
colleagues Urban aide his co-pi together with me and then Mike Fisher he's a 1:11:35
former CTO of etsy he's our double Alum and he has been working with us as a technical advisor also our wonderful
1:11:43
collaborators from University of Pittsburgh Ali is here Morgan Frank faculty member at University of
1:11:50
Pittsburgh and Sarah bana Chapman University so these are our team 1:11:55
collaborators very briefly the project that we worked on is a part of a larger 1:12:01
project that we are doing um you know responsible digital innovation in the ler space we just completed a report
1:12:09
fairly comprehensive report on the design patterns and principles of universal responsible Universal ler
1:12:17
ecosystem for those who are interested in reading the report I'll be happy to share the report with you via email as
1:12:25
part of this project we thought that it'll be useful to uh demonstrate the 1:12:30
capability leveraging the large language model that our collaborators have built 1:12:36
automatically issuing w3c compliant verifiable micro credentials turning a 1:12:44
single credential at the transcript and diploma level into series of micro 1:12:49
credentials that our former student has accomplished so it has several key components the 1:12:56
first one is ingesting the uh the unstructured PDF from transcript turning 1:13:03
into a series of you know identifying courses and grades and then we pull down 1:13:09
the syllable eye that of the courses that are included in the transcript then 1:13:14
using the uh the large language model to ident to extract a skill Vector with the 1:13:22
weight that of the student and then using our heuristic algorithm to identify or infer 1:13:29
the skill sets that the student might have accomplished during that program and 1:13:36
then we use velocity networks credential engine to produce verifiable credentials 1:13:42
and then issue them into their wallet so I'm going to turn it over to First the 1:13:48
video created by Ali you're going to his hear his voice is here in the room and then 1:13:56
after that ishan will take over and then explain our heuristic algorithm and then show you the demo video to conclude the
1:14:03
presentation in this work we developed the syllabus 200 pipeline as a framework that enables 1:14:09
policy makers and decision makers to extract only the skills from core syllabi this work is a collaboration
1:14:16
between the University of Pittsburgh Chapman University and Teachers College Columbia University
1:14:23
in order to investigate how college education prepares the students for the labor market we developed a framework to
1:14:30
extract the skills being taught at higher education institutions their curriculum and map them to what
1:14:37
activities required by the labor market the framework receives a road syllabus 1:14:42
as input and breaks it into sentences to keep only the course description related sentences for the next steps then the
1:14:50
framework utilizes language models to compare the semantic information in a course description to the semantic
1:14:57
information encoded in the skill Tech summary it is worth mentioning that in 1:15:02
this work we use the onet detail work activities as their skills taxonomy
1:15:07
which can be replaced by other skin taxonomies to meet a specific damaged needs finally the framework calculates
1:15:15
the pairwise similarity between each sentence and a scheme and aggregation scores for each skill across different
1:15:22
sentences as a result a wrong core syllabus is transformed into a vector
1:15:28
representing its similarities scores with each of their skills
1:15:36
all right so in a sense what happens is when we take in a student transcript in the PDF format it goes through our large
1:15:43
language model which we have in our back end takes in are the coarsely by which you also have in our backend and kind of
1:15:48
gives us a vector of the skills along with this course which we have that if you take took this course out of these
1:15:54
skills these are the top skills you're lucky to have achieved after completing this course but as a student you don't
1:16:00
take just one course you take a lot of courses I don't know probably hundreds of courses they are like on different
1:16:06
levels different course credits different grades and each course reinforces somewhat what you learned in
1:16:12
the past courses so how do you kind of aggregate all of these factors and give that hey you complete these 10 courses
1:16:20
out of your undergrad degree this is the all after committing all these courses these are the skills which you are most
1:16:26
likely to have achieved so cons there are a lot of factors which go into that play but some of the factors which we considered for our purposes was the
1:16:34
course credits like if you take like three point or 4.0 course that's a factor which we considered what's the
1:16:40
grade you achieved and one other factor which we also considered was how a future course reinforces the skills you
1:16:47
achieved by completing the past courses for example you know you took CS courses 10 CS courses and you took just two math
1:16:54
courses then you're more likely to be good at technical skills like Java or python as compared to calculus or
1:16:59
statistics so considering these this heuristic algorithm what happens is we take that scoring uh the score Vector
1:17:06
multiply it with our course credit that's a first step to consider the course credit for each of the courses and then we include the great factor
1:17:14
which is also available on my screen which is if you come took an a grade if you've got an A grade this is the the
1:17:20
multiplication factor in b c and d so that kind of reinforces that depending upon your performances for the grade for
1:17:26
the course this is you know the level of the skills you achieve and finally we aggregate the sum of all these skill
1:17:32
vectors for all of those individual courses together which kind of brings me to my same point which I was making earlier that reinforcing how the past
1:17:40
courses have and the new courses you're taking Aggregates and reinforces your skill learning through your uh undergrad
1:17:46
or whatever degree you're taking it so this is a kind of the heuristic algorithm we take some of the factors which we consider there are a lot of
1:17:51
more factors outside of this Factor but kind of this gives an idea that using this heuristic algorithm we can go down
1:17:58
that after you took this level of courses uh and the course level which is in there a large language model is able
1:18:04
to provide the skill vectors for us and then finally a curated list of the skills based upon this heuristic
1:18:10
algorithm so now I'm gonna head over to our demo which is how this entire thing comes 1:18:15
into the play uh for our application hi everyone welcome to the demo of skill 1:18:22
certify AI I'm ishan Gupta Project Lead for this amazing Innovative Solution by 1:18:28
University skill certify AI it is a natural language processing framework model it 1:18:33
is Packet along with a blockchain based utility layer to help map co-service by 1:18:39
into skill competencies for students and then generate micro credentials recognizing those skills achieved
1:18:47
so let's see how this goes the first step is the account registration by the students when they 1:18:53
use their email address provided by University and upload their unofficial student transcript 1:18:59
the next student gets to review the list of all the courses they completed as well as the grades they accomplished for
1:19:06
them moreover the students can review the syllabus for all the courses the courses they took and completed as part
1:19:13
of their academic Journey thirdly using our natural language processing model the students can review
1:19:21
the list of achievable skills for all those individual courses they
1:19:26
accomplished as part of their academic Journey once that's done the students get to see a curated list of all the
1:19:33
skills they might have accomplished by completing the list of all the courses they completed as part of their Academy
1:19:40
Journey this is done using an heuristic algorithm which I'll talk later in this video but I once the students can see
1:19:49
that curated list the students can self-validate claiming whether they actually feel they have accomplished
1:19:55
those skills or not after this self-validation step the students can finally generate blockchain anchor
1:20:01
digital credentials for all those individual self validated skills and can share those credentials with anyone who
1:20:08
they feel like so let's see how this goes [Music]
1:20:15
the first step is a contact registration so I'm going to put in my name my email 1:20:21
address [Music] my username
1:20:26
a password [Music]
1:20:33
once that's done it asks me to pick the choice of my degree since I was an undergrad student I'm going to choose
1:20:39
undergrad and now it asks me to upload my student transcript
1:20:44
I already had my University student transcript in its PDF format with all the kind of courses with their grades
1:20:49
listed over here so I'm going to upload that right over here so choose file 1:20:57
where is my transcript transcript you can upload bias
1:21:02
transcript over here and then click on Academy progress this initiates a data 1:21:08
pipeline process for where the data from the transcript is extracted and loaded into our sqlite server in the packet
1:21:14
first that's done the students can review the list of all the academic courses they completed along with the
1:21:20
grades over here they can also see the syllabus for all the courses uh they 1:21:25
took at their academic Institution so once that's done I'm going to click
1:21:31
on review kind of progress this is going to pass in all the course syllabi into our natural language processing model which generates the scores for all the
1:21:39
possible skills students might have achieved so over here you can see that if I took 1:21:45
an econ course based upon the course thereby the list of achievable skills for this course is going to include
1:21:50
estimation cost of goods and services determining pricing or monetary policies calculating costs of goods and services
1:21:56
and so all those skills like that what if I took and computer security course or a computer security course I'm
1:22:03
probably going to achieve skills in analysis security of systems Network or data Implement security measures for
1:22:09
computer information systems monitor the security of digital information and all those skills 1:22:14
what if I took another uh can uh or probably CS course let's say 345.
1:22:21
so for a program language Concepts class these are the list of skills which I might have achieved so these are the
1:22:27
individual skill set of skills I might have achieved by completing these individual courses but now I'm more
1:22:32
interested in a curated list of skills I have achieved by completing all these courses at my Academy institution so for
1:22:40
that I'm going to click over here this initiates the heuristic algorithm microservice on our backend which takes
1:22:46
in all these uh scores for all these individual skills and then
1:22:51
consider factors such as grades for the courses I accomplished uh the course credit for each courses and another
1:22:59
critical Factor we uh this algorithm takes in is the number of occurrence of 1:23:05
the skill across multiple courses what I mean by that is let's say a skill has a 1:23:10
high score across seven uh different courses then that skill is likely to be 1:23:15
achieved by me as compared to a skill which had a high score across just three 1:23:20
courses so that that the larger the overlap uh the skill has the higher the
1:23:27
chances it has been achieved by me so consider heuristic algorithm it this 1:23:32
is the list of which of skills which I might have accomplished after completing 1:23:37
all those courses over here it lists top 20 skills uh uh 20 is a higher parameter that could be 1:23:44
tuned depending upon the needs of this application but at this step students are asked to self validate whether
1:23:52
these skills if they believe this achieve this skills or not so since I took a lot of security courses I believe
1:23:57
I have achieved that skill since um you know I took a lot of projects and developed reports I think I
1:24:04
have achieved this Kelly as well given I have taken third level of calculus course I'm pretty sure I'm good with
1:24:10
Statistics and mathematical courses um since I've taken a lot of Technology 1:24:15
based courses I'm gonna go with this skill as well so I can I can validate all of those skills but for now for the
1:24:21
demo purposes I'm just going to pick the top four which I feel I have accomplished so once 1:24:27
I'm done I click on validate skills and initiate the general potentials for these four skills so this 1:24:33
hits our microservice in the back end which uh takes in all these the self-validated skills and starts
1:24:39
generating micro credentials for that uh just give it a minute it's a few minutes before it starts the 1:24:45
process
1:24:57
okay there we go our process our certifies are generated a Json file is also generated which shows the
1:25:04
information about the vendor user ID uh you know vendor endpoint and all the 1:25:10
information regarding which might be concerned as per our certificate issuer [Music] 1:25:17
what is important for for the students is this QR code which is generated which the students can scan using this
1:25:22
velocity Network Foundation app and can download and claim all the micro credentials which were generated after
1:25:28
the last step so this is how the entire application flows starting from the University 1:25:34
student transcript in a PDF format to finally blockchain anchor digital credentials in 1:25:40
Json format we saw this from the undergraduate student Persona I have another Persona 1:25:46
prepared for us just to demonstrate how right so that was part of our demo and I'm going to hand over to Professor you
1:25:53
and he's going there all right well that was a pretty awesome and we are very happy that we were able to demonstrate
1:25:58
uh the capability uh and the next step is uh we the the skill sets that we are 1:26:06
relying on is on its skill set but we are working with our partners to expand 1:26:11
those skill sets from other databases as well and we are trying to uh use this 1:26:18
technology to apply in certain industry sectors so thank you for your attention 1:26:24
thanks [Applause]
1:26:38
all right if you haven't taken a moment take a moment and say something to somebody next to you about something you've seen so far today
1:26:46
just something we gotta we gotta move you know you can't sit too long
1:27:07
it's like we had a sad a sad person at a table all alone
1:27:23
all right so it's a good chance to remind everybody that after
1:27:30
there will be a networking uh two-hour networking block for you to engage in conversations with 1:27:36
your peers just as you are doing now
1:27:53
thank you
1:27:58
foreign
1:28:04
we're going to bring it back up front please
1:28:10
so I spent my first uh three years as a first grade classroom teacher and one of 1:28:17
my favorite things was the quiet coyote does anyone remember the quiet coyote yeah there's the quiet coyote this guy
1:28:22
remembers there are a lot of those skills that translate and I wouldn't actually know 1:28:28
how to like capture them into a credential but I would love to have a conversation with one of the tools we've seen so far
1:28:35
to try to put that into a package to bring that forth with with my experience and to help me find 1:28:41
where that applies where is that a useful skill so we'll have a networking break after 1:28:49
uh this this next set of groups we're going to transition now into a new use case there'll be a short networking
1:28:56
break after this and then at the end of the day just a reminder all the teams that are up here they will have tables a
1:29:02
table for you to engage in conversation about their project specifically there'll be a little table placards up
1:29:09
on their respective areas but that two-hour block at the end is meant for you to be able to go up and engage in a
1:29:15
deeper conversation ask more questions about that work uh next up our our next use case 1:29:23
uh is built around uh the military pers uh military personnel
1:29:29
uh and this idea of uh the soldier sort of loosely based um which will be this 1:29:34
next slide that's coming up a reminder how these were made these weren't just like plucked out of the air and just thrown on paper uh these were
1:29:41
actually uh you'll notice that on this next slide they're quotes these were lines that were taken directly from
1:29:47
individuals that these project teams had conversations with about their experience about their credentials about
1:29:55
their uh job seeking journey and what that looks like so project teams went 1:30:03
through a series of of activities as part of the experience you project to
1:30:08
actually map out what some of the unintended consequences of these Technologies might be on the individuals
1:30:14
that are using them project teams were also asked to map out some of the uh to map out some equity
1:30:21
and bias considerations around their data sets and around the solutions where 1:30:26
is your data coming from how has that data collected how is it aggregated and how is it how is it informing the
1:30:32
systems you're creating and the solutions You're Building the idea of a large language model is 1:30:38
really great but where is that large language model collecting information and how is it 1:30:45
informing what you're making so as we mentioned some of these Solutions impact more than one use case
1:30:52
and this next group is a great example of that somebody who was in the military is transitioning to a civilian role
1:30:58
probably has to go through a lot of the same certification exams if you're going to be a teacher there is a set of exams
1:31:04
you have to take and you have to pass those education testing services is the 1:31:10
world's largest private non-profit education testing and assessment organization ETS develops various standardized tests
1:31:17
for K-12 higher ed and the workforce ETS owns International tests including 1:31:22
the teacher test of English as a foreign language the test of English for International Communication and The
1:31:28
Graduate record examination better known as the GRE the TOEFL and 1:31:36
the toeic these tests have been administered in
1:31:42
over 180 countries and at over 9 000 locations worldwide the Project Lead for 1:31:48
this project from ETS is Guang Ming Ling and he is a research director for the 1:31:54
center of education and career development research and measurement science area at ETS Guan Ming's current
1:32:01
research focuses on the assessment and training learning of critical competencies such as critical thinking
1:32:07
and collaborative problem solving in higher education and Workforce settings 1:32:13
he also works on student learning outcomes assessments and factors related to the validity reliability and fairness
1:32:19
and Equity issues in the education and Personnel selections Dr Lang holds a U.S 1:32:25
patent on the technology in relation to using an automated scoring system to improve scoring accuracy and efficiency
1:32:31
of speaking responses in English his research has appeared in over 100 presentations at International and
1:32:37
national conferences and over 40 papers in peer-reviewed journals we'll turn the stage over to Guan Ming Ling with
1:32:44
education testing services [Applause]
Use Case Group 2: Military Veteran
1:32:59
thank you thank you for the introduction and nice to see you guys and thanks for coming 1:33:05
um as a um Colleen just to introduce you know uh we have a team that was 1:33:11
assembled around March and uh it's a little bit late in the game and the goal 1:33:17
there is try to follow the uh requirement and to try out ETS approach
1:33:24
to map the contributed to the experiencing efforts
1:33:30
so over overall going we try to create and present a proof of a concept demo in 1:33:37
which we try to help incumbent workers I will explain a little bit later learning 1:33:42
and working records into a skill profile based on the Open Standards skills framework 1:33:50
I think this is a little bit different from what Colleen just ensures there are some confusions here we did start with
1:33:57
uh uh one of the user cases from the
1:34:03
military background and but we didn't wasn't able to work out because of lack 1:34:09
of information so today instead I'll present the case that we were able to go 1:34:14
through foreign workcase
1:34:21
um so the uh we're going to have a brief introduction about the um this category 1:34:28
of incumbent workers profile and then we can have uh the introduction of 1:34:34
different elements that we have go through following the uh the charts that introduced earlier and we also try to
1:34:42
emphasize the uh how we fulfill the requirements and
1:34:50
also we can share with everyone a demo clip video clip where we show how this 1:34:56
tryout works and we're also going to check a little bit about the next steps 1:35:02
after the experience use stage so how this works and we started with 1:35:09
multiple considerations besides that we need to have the input from the 1:35:14
individual workers and um we need to find identify an Open
1:35:20
Standards of skills framework we also need to have equity and bias
1:35:27
issues considered as well as the conformance test when we try to build the system all these considerations was
1:35:34
added to the AI based algorithm and from there we try to extract the
1:35:41
skills profile for individuals and we also try to cross check you know whether 1:35:48
AI is doing an okay job or reasonable job to identify the skills and also 1:35:53
identify related evidence in support of individual skills and after that we we were able to 1:36:01
extract the two sets of outcomes one is a human view outcome of the skills profile so that individual users can use
1:36:09
them and check is see if everything was captured from their understanding but 1:36:14
also to share with end users for example possible employers and other use cases 1:36:21
and the second part is we also we generate the machine view which is a 1:36:26
Json file following the requirements which can be utilized in different next 1:36:32
steps of utilizing the experiencing results
1:36:38
um in terms of the unstructed data input in this case we're
1:36:43
mainly focusing on initial stage the cover letter and individual CV as two 1:36:49
formats in PDF formats we did as I said earlier you know we did interview the uh 1:36:56
um the other military case and also three simulated cases as well but for this demo purpose we're only going to
1:37:02
introduce the results from the case four which is a live case introduced from The Experience use team
1:37:10
and we took the advantage of Open Standards uh in this case the ownet 1:37:16
skills we started with only top five skills for this demo purpose of course you know after this if folks are
1:37:23
interested in we can easily extend to uh any number of skills that using this AI 1:37:30
engine we also explored the credential ending well they have tens of thousands of 1:37:36
credentials which is taxed to different skills but after several weeks of effort 1:37:42
we were not able to come up with a clean defined structure of skills so we we 1:37:47
didn't move in that direction um so in terms of the evaluation of AI
1:37:54
extracted skills profile we evaluate the agreement between the
1:38:00
machine extractor skills but also identify the evidence against those from the skills 1:38:06
experts at ETS in most cases we found that AI
1:38:12
identified evidence include includes those those who was identified by human 1:38:19
but in terms of the skills presents they seems mostly overlapped with each other 1:38:25
very well and the additional evidence identified the AI image process was also 1:38:33
seems reasonable based on the single case we evaluated so here's the comparison results between
1:38:41
the human and the Machine identified skills this seems to be for this a single case that we were able to
1:38:48
experiment and they seemed to be perfectly agreed to each other um however 1:38:54
um we feel this is only one case and this might be too good to be true in the 1:39:01
General application for other cases so we would put some cultures there that you know we
1:39:08
need to do additional evaluation for broader user cases in the same scenario 1:39:14
this is an example of the interface that we put out there I know it's probably 1:39:21
hard to read we have a video clip towards the end of the presentation
1:39:27
and uh in terms of the equity and bias considerations there have been many discussion in the industry in the AI
1:39:34
about the possible unintended consequences of using AI
1:39:40
from our user case you know we manually try to investigate the implicit and 1:39:46
Expedition the bias in terms of the development and use of algorithms for 1:39:51
competence translation here uh for example we want to make sure users in 1:39:56
this case income modern workers know and understand uh the input and output from the AI and we want to focus on the
1:40:04
assets of individuals that already mastered rather than what they don't
1:40:10
have yet and also we want to ensure users knows what to do and what are the actionable 1:40:17
information from the output of the AI and we also thought about the next steps 1:40:24
you know we could use some sentiment language analysis to examine the output 1:40:30
of AI and to see how that could be making it more Humane and readable and there might be multiple ways to
1:40:37
present results as well different from what we're going to show later on as a result of this experiment
1:40:43
this is mainly talk about the conformance test where we met we were
1:40:49
able to meet all the requirements and just to make sure I highlighted this is 1:40:55
obv3 It's a self-skill assertion reported by individuals based on their
1:41:01
CVO cover letter it's not verified by any third party yet
1:41:07
and this is the conformance test and [Music]
1:41:12
let's see um in terms of the expandability as part of the requirement
1:41:18
um I think you know we we have enough evidence at this point uh the AI seems 1:41:24
reasonable for now that uh to generate or to map individuals skills or
1:41:30
experience into this skill profile in some cases for example a lot of
1:41:36
incumbent workers even the military background workers they may not have 1:41:42
enough time or experience prepared for their CV in those cases when the CVO cover letter was read or processed using
1:41:49
AI they may not generate as good example as we are presenting today so I think that's a caution that we want
1:41:55
to make a note here and finally this is a video demo we want
1:42:00
to share with everyone how we designed the process so
1:42:07
ETS skills profile system will show houses tool can help extracting valuable skills and information from user
1:42:14
documents for example resumes cover letters and certificates to start upload the documents you want
1:42:20
to analyze you can click on the upload section or drag and drop directly from your computer 1:42:29
so next step is to select a skill framework option that you prefer we
1:42:34
currently have options for our net and greater your own framework [Music] 1:42:40
before submitting you can confirm the container real documents in the drop down box below 1:42:50
now that we are ready to start analyzing your document the processing time may 1:42:56
vary depending on the size and complexity of your documents after the analysis is complete you can view the
1:43:03
resource in the right and drop the Box corresponding to your selected Frameworks 1:43:08
here you will find a comprehensive list of skills detected along with the evidence in the documents
1:43:15
Note 2 is perfect and it's always a good idea to review the content to ensure its 1:43:20
accuracy
1:43:27
additionally using the feedback box the user can let us know about any discrepancy suggestions or Improvement
1:43:34
to our tool and that wraps up the demo of the etsk profile system
1:43:41
so as a result we also were able to go back to the user case
1:43:48
um that we initially talked with and shared with her the results we were able 1:43:53
to generate and we want to seek her initial feedback and it turned out she 1:43:59
was really uh surprised and happy to see the results and one comment she made was 1:44:04
that you know it'd be great and if she can use this information when she tried to look for New Opportunities and share
1:44:11
with future employers and she specifically knows several of the skills
1:44:16
identified from the CV and the cover letter as instructed by the CV she
1:44:22
mentioned that there are multiple friends and co-workers have mentioned this to her that she has those skills
1:44:28
but in the CV it was not expressly called out she feel like you know if someone like ETS could confirm these
1:44:36
skills actually was present from her experience they seems such a huge Plus for her to to as a use as a leverage to
1:44:44
seek additional opportunities um and uh last but not least you know I
1:44:50
want to thank this is definitely a teamwork uh from ETS and uh we have
1:44:56
several colleagues and one of them are here um Diego uh and we also have Blair Carol 1:45:04
uh Andrew keys and uh Jenga and Lydia and finally I want to thank
1:45:09
organizations who make this possible for today's event and also especially the 1:45:15
educational Design Lab team Taylor Colleen and Phil and thank you for all 1:45:21
your uh endless support and very generous and also the one user case
1:45:26
which I don't want to mention them for confidentiality issue and thank you guys 1:45:31
and we're going to have a table in the back and we're happy to discuss first if you have any interest or questions
1:45:38
thanks thank you
1:45:54
all right for our next group uh we have a collaborative effort between astrum uh and national student
1:46:01
Clearinghouse so ask from you translates educational experiences into Economic Opportunity they're on a mission to
1:46:08
quantify the return on education Investments for Learners education
1:46:13
providers and employers ask from you helps institutions measure the value created for incoming and returning
1:46:20
students while assisting them in securing industry Partnerships that lead students seamlessly into high demand
1:46:26
career Pathways institutions partner with astrom u to drive enrollment and increase alumni and corporate engagement
1:46:32
while extending economic Mobility opportunities inclusively to all learners 1:46:38
national student Clearinghouse very familiar group to me as an educator is a non-profit formed in 1993 and it's
1:46:45
a trusted source and leading provider of higher education verification and electronic education record exchanges
1:46:52
besides working with nearly 3 600 post-secondary institutions the Clearinghouse also provides thousands of
1:46:58
high schools and districts with continuing Collegiate enrollment progression and completion statistics on
1:47:04
their alumni the project leads for these two groups are cavorn Anderson a Solutions 1:47:11
architect from astrom uh and he works with Astra mu as an architect with a 1:47:17
drive for problem solving creative thinking and interpreting data to make optimal decisions as an arc a Solutions
1:47:23
architect he faces engagements throughout pre-sales and post-sales activities he applies a strategic
1:47:29
mindset and brings his technical expertise Within These opportunities to help bridge the gap between sales and
1:47:35
execution from national student Clearinghouse we have Krupa arhia she is a specialist 1:47:42
business analyst and an Innovative creative individual with a passion for business analytics she's demonstrated
1:47:49
experience in designing and delivering highly effective business intelligence Solutions she's a self-motivated learner
1:47:55
with skilled in gathering and synthesizing data into actionable insights to enhance organizational 1:48:01
performance proficient in interpreting and implementing business requirements technical specs and functional
1:48:07
requirements well versed in design developing and implementing user interfaces and 1:48:12
database objects while maintaining excellent written verbal communication skills both within the corporate and the
1:48:19
technology communities these are our next groups astrom U and national student Clearinghouse
1:48:25
[Applause] [Music]
Use Case Group 2: Military Veteran
1:48:37
thank you hello good afternoon thank you so very much Colin I really appreciate that uh I 1:48:44
want to just say thank you to the Chamber of Commerce education Design Lab in the Bill and Melinda Gates Foundation
1:48:49
for opportunity to be here today to talk to you about what astromew and the national student clearance house has
1:48:55
been up to these past couple months and what our research has been about the past couple years so
1:49:00
I'm Kev Warren I'm a solution architas from you we are at Bellevue Washington based Ai and data services company and
1:49:06
I'll let my colleague go ahead and introduce yourself good afternoon everyone I'm Krupa audio I'm the product owner at National Clearinghouse a
1:49:14
Education and Research non-profit based out of Herndon Virginia
1:49:19
thank you Krupa so create an efficiency transparency and opportunity for Learners are some of the key drivers
1:49:26
behind our partnership with the national student Clearing House and lers will be one of a very valuable
1:49:32
asset towards achieving those objectives the continued research and investment into the adoption and interoperability
1:49:39
of lers um will can have a significant impact
1:49:44
and will have a significant impact for the outcome of those Learners and especially in our case for military
1:49:50
veterans of our of our nation so why does this even matter so I know that there's a lot of 1:49:56
tremendous work today for from organ from government organizations from 1:50:01
non-profits and from industry organizations as well to help soldiers with their transition from serving in
1:50:07
the military to to the civilian life but we can all agree and there is evidence 1:50:12
that shows that there is much more work to be done for example there are over 200 000 1:50:19
soldiers that leave the military every single year and only a fraction of those 1:50:24
soldiers today are getting a job other than their first year however within the 1:50:29
next one to two years a lot of those soldiers actually end up transitioning from their first job our thesis is that
1:50:38
a well-designed ler solution well can and will help veterans get better get
1:50:45
better jobs faster so let's take a minute and let's talk about what a well-designed ler solution looks like
1:50:51
and how that would work for a real person consider Jonathan he spent his entire 1:50:57
career within the Army and now in E7 he needs to transition out of the military service a lot earlier than what he's
1:51:03
expected but as he's coming to realize just like a lot of people these past couple months 1:51:09
the job market is very complicated and is very unpredictable
1:51:14
the stress of finding a new job can definitely cause a lot of anxiety especially in Jonathan's case because
1:51:22
there are a lot of companies out there just like initech who have committed themselves to hiring a lot of veterans
1:51:28
across all of their jobs but intersect themselves they're very unfamiliar with military documents in 1:51:35
Army speak but thankfully for Jonathan ask from you and the national student Clearing House 1:51:41
ler solution will help him to translate his military experience into skills that 1:51:47
are presented in a portable format that is on a Level Playing Field with civilian experience but before we do
1:51:54
jump into the demo and show you the technical aspect of it I like to address some of the ways that
1:51:59
our team has been working on ensuring that we can combat bias within our AI 1:52:04
models so starting off depth and breadth the more detailed and
1:52:11
diverse the data that you train your model against the better the results that you actually yield at ash from you
1:52:17
we've actually trained our AI model against millions of data points from over 10 years worth of army records that
1:52:24
are available this has allowed our AI model to understand the context of each 1:52:29
of those records and extract the skills and abilities within a degree of
1:52:34
confidence without relying on simple keyword matching next The Secret of data science and data
1:52:42
management is that all data is dirty so to make sure that all of our research 1:52:47
and all of our assumptions are correct we have worked with former military advisors and 1:52:53
experts in the space to help us tag and tag and model our training data and the 1:52:59
schema that we end up actually creating is based on the successful outcomes of soldiers into the into into civilian
1:53:07
speak so this was done so that the model does not exclude or favor specific skills over another and
1:53:15
ensure that the results that we do produce at the very end is from a civilian's perspective and lastly
1:53:21
we understand that learning happens everywhere just because you didn't finish every 1:53:26
aspect of your degree or your certification path it doesn't mean that you didn't learn something real specific from the classes that you took so in
1:53:34
order to not be exclusive of that progress we've actually trained our AI model to 1:53:41
it to against those individual classes themselves to so that you can extract the Learned skills and abilities
1:53:48
so let's actually get into the demo and one thing to clarify the individ the
1:53:54
information that you will see through this throughout this video is based on someone's real actual joint service transcript and service record brief so
1:54:02
this is a real this is a real document that we were able to parse but we have it has been de-identified and anonymized
1:54:08
for the privacy and protection of the soldier themselves so let's start it up 1:54:13
Jonathan has both his joint Services transcript and his service record brief these are documents provided in a
1:54:19
summary of all this training service experience and his accolades using a simple UI Jonathan can upload
1:54:26
both of these records to them be processed by astrom's translation engine which then provides him with a few
1:54:32
different output options astrom use AI then extracts the raw data
1:54:38
which is organized labeled and structured into a Json file this data includes Jonathan's basic personal
1:54:44
information as well as his job MOS details and training and education that 1:54:50
he's completed while in service his other output option uses natural
1:54:55
language processing to extract skills information from the description and adds additional skills data from both
1:55:01
proprietary and public sources such as the American Council on education 1:55:12
as an additional feature option the AI translation engine uses cluster and 1:55:17
topic modeling to dynamically categorize all the skills from both his experiences and education 1:55:27
all of the structured data is available for Jonathan to download as a Json file 1:55:33
now that Jonathan understands his skills and has a digital breakdown of both of his verifiable records we can transform
1:55:39
his data in multiple ways for the purposes of screen Sue we can generate an open bad version 3 compliant
1:55:45
credential that could be downloaded and ported into any digital repository this is an example of generated a credential
1:55:52
for one of Jonathan's MOS codes this credential data can be exported as 1:55:58
a Json ld5
1:56:07
which we can then validate see if it meets a requirement of an open
1:56:12
bag version 3 Json LD format
1:56:22
very simple straight to the point Jonathan now has a digital credential
1:56:27
for one of his occupational Specialties let's take a look at the clearing House's My Hub and see what Jonathan can
1:56:33
do with his now self-asserted credential thank you Maya was created by the 1:56:39
Clearing House about four years ago and we started with John Hopkins University since then we've grown to serve about
1:56:47
150 schools and are constantly adding more My Hub is a data aggregating 1:56:54
platform where a student can view all their degree data from all different organizations 1:57:01
their credentials the enrollment data and from all participating organizations 1:57:08
students can download their certificates such as enrollment certificates good student discount certificates and
1:57:15
download their Clearinghouse transcripts from within the platform additionally we 1:57:20
have also integrated with third parties such as Paradigm credly and badger 1:57:25
for this for the purpose of this demo we've used badger
1:57:31
so we've created a real with using real skills extracted from the joint service transcript by Astra mu we've used Badger
1:57:39
uh to create a test is sure and test badges to import into the into
1:57:46
credentials into my Hub why we used a test issuer and test badges for
1:57:53
experience you this was done using a real service real joint service
1:57:58
transcript and real skills were extracted so one so now that once Jonathan logs 1:58:05
back into my Hub he would see his service record briefs his joint service 1:58:10
transcripts translated into badges along with his degrees and enrollment data this paints a more holistic picture of
1:58:18
Jonathan's Learning Journey and skills it includes both his civil civil and
1:58:24
Military data in the same place myhub is also currently working on additional 1:58:30
Partnerships to allow Jonathan to share these degrees and credential data onto
1:58:35
an employment Marketplace to get hired based on his skills and data and hopefully Jonathan can find an
1:58:42
appropriate job quickly and give credibility to his skills
1:58:53
active thank you so much so you've all seen a solution Jonathan has what he 1:59:00
needs and this is only the beginning the same AI capabilities being used today for astrom use free to use tool
1:59:07
for Army soldiers to help them understand what life what transitioning out of the military service looks like
1:59:13
using those very same documents that we showed you before today's ler can help Jonathan but there's still over 200 000
1:59:20
service members that are leaving the military every single year so what if what is the potential impact
1:59:26
if we were to start shipping at that 12 of soldiers that it takes them a whole year to find their very first job
1:59:33
what if we could help reduce the time from one year to just six months of them 1:59:39
find the job for only 20 of those veterans for veterans that out that end up making 1:59:46
an average around seventy thousand dollars a year reducing that time from a year to six months could actually could
1:59:53
add an estimated 116 million dollars back into the pockets of those veterans 2:00:00
and reduce their need for VA benefits payout by an estimated 84 million
2:00:05
dollars so what about the ones that are leaving not 44 percent
2:00:12
given that it takes on average about five months for someone to find a new 2:00:17
job if our solution could help reduce churn and eliminate the need for just 10 2:00:24
percent of those veterans to find a new job we could save this group an
2:00:30
estimated 242 million dollars in potentially and
2:00:36
otherwise lost wages and it and given it does cost an
2:00:41
employer on on the very low end at least 90 percent of their employ of that 2:00:47
employee's annual salary to replace that employee's productivity employers of 2:00:53
themselves could save an estimated 523 million dollars each and every year 2:01:00
a civilian over a billion dollars worth of value back into the economies of our 2:01:05
communities because Jonathan has now found a better aligned job in a very shorter amount of 2:01:12
time he's much more satisfied within the square and using he and he's now using 2:01:18
his skills in a much more effective way and the economy itself is stronger for it thank you 2:01:23
thank you
2:01:46
one of the things I've Loved about this project is being able to get to know these project leads because they they've
2:01:53
invested a significant amount of their time and mental effort to to get to this 2:01:58
point to be able to to step up on stage and just make it look like this is part of what they do every day a lot of these
2:02:04
teams and these lead leaders had to make some serious considerations about their product roadmaps about their how much
2:02:10
time they want to commit during you know their typical work week think about if a new project got slotted in on you and it
2:02:16
was something that you initiated or somebody else told you you were going to do we've had some we've been some pretty
2:02:22
fortunate we've been pretty fortunate to have some awesome Champions step up and and really lead in this space
2:02:29
our next group is oh a housekeeping note that I'm supposed to tell you you have a 2:02:34
badge if you are attending the T3 meeting tomorrow make sure you keep that badge and use it again tomorrow so that
2:02:43
was one thing I needed to pass along to you our next group is is edgeworks and Edge
2:02:49
work supplies Advanced Technologies to deliver solutions to the biggest challenges in human performance lifelong
2:02:55
learning and career planning Edge work Solutions have consistently pushed the envelope in human first applications of
2:03:01
AI edgeworks and partner organizations have worked together to present A New data-driven Perspective on learning and
2:03:09
careers now Alan LaFleur and Robbie Robson were two of the project leads who sort of 2:03:14
were collaborating on this effort up until the demonstration on the back end 2:03:20
and presenting today is Fritz Ray he is the chief technology officer of
2:03:25
edgeworks Corporation with 20 years of experience developing software for use in academic government commercial and
2:03:32
open source environments including applications for Aviation financial and patent areas he leads software
2:03:39
development and Edge works and contributes to IEEE and other standards efforts around learning and is the lead
2:03:45
contributor of the open source Cass project so I'll hand over the stage to 2:03:51
Fritz Ray [Applause]
2:03:56
[Music] thank you
Use Case Group 2: Military Veteran
2:04:01
[Music] thank you
2:04:09
hi so uh my talk today is going to be much more focused on the use case 2:04:16
um so we're looking at an individual sailor here let me give you a little bit of background on edgeworks Corporation
2:04:21
and then we'll get into it so we're a small business established 2001 we've been supporting the US government and
2:04:28
DOD and doing research development in the training and education space for a 2:04:35
couple of decades now we have projects with Army Navy Air Force DARPA NSF and 2:04:41
have been maintaining a portfolio to include AI large language models all the 2:04:47
way back to latent semantic analysis and playing dairishly allocation if there's any old school language modeling folks
2:04:53
out there and that's Pursuit you know continued into what we're doing today 2:04:59
with large language models Etc
2:05:04
so we're gonna be talking about Susan Malik who's a transitioning Navy sailor they are an os3 and they've been
2:05:11
recently deployed on the ddg 86 Shoop sailor lingo is going to kind of feature 2:05:17
throughout this because the world in which they inhabit is very different than ours over here on the the
2:05:23
commercial side so when Susan's thinking about getting
2:05:28
out of the Navy and transitioning to civilian life she has resources that are available to her there's programs like
2:05:35
DOD cool and Mill gears that help her identify based on her rating that is her 2:05:41
occupation what she may be good at out in the commercial sector this uses onet 2:05:48
data along with a bunch of other data but it doesn't really articulate too much past the occupational level and
2:05:55
some courses she also has various funding
2:06:00
opportunities through something called DOD skill bridge where she can get some of her time paid for as she transitions
2:06:06
into an organization that works with the the skill Bridge
2:06:12
and so that's been there and she has ways of of transferring her
2:06:17
credentials out but typically she has to go and take the credentials on the civilian side again so if there's a a
2:06:24
degree she would need to go basically take a large assessment for that degree or or that credential
2:06:32
there's also programs like US map which helps with apprenticeship hours requirements and other sort of
2:06:40
apprenticeship type programs for Blue Collar work primarily but that's around 2:06:46
and and able to assist her but there's not really a system out there for
2:06:51
um for translating her credentials directly from the military into civilian life and 2:06:57
on the left here um is what we all know and understand something I call the learner's Journey 2:07:03
where we start with aptitude we move on to Education and Training certification qualification practice and performance
2:07:09
this is the stack that gets you from where you are to where you need to be to be productive and on the right side is a
2:07:15
sailor's journey and it's very different they start off with an ASVAB they go to a school C school nec's pqs they take
2:07:22
evolutions and drills to get their ships out to sea and then they deploy on the left you have a resume and a
2:07:29
transcript on the right you have things like the US map record and a training jacket and so continuously We're
2:07:35
translating between these different languages for the different services
2:07:40
so we're going to go into a demonstration now and I'm going to step
2:07:45
back and we'll see what this journey entails
2:07:51
we're going to go through an example of the Navy service branch of the US 2:07:57
Military and so here we start with a the upload of a training jacket which is a 2:08:03
document that lists out the courses credentials qualification and
2:08:09
deployment of in this case Susan Malik so upload that we hit next that's
2:08:16
extracted using edgeworks document parser some of the things that are happening while this is being extracted
2:08:21
so let's talk first a little bit about addressing bias within within Ai and and 2:08:27
specifically the approaches we use which are large language models primarily and concept models and um so large language
2:08:34
models as public opacity at all in 2016
2:08:40
um found out our largely biased based on the sort of statistical mean of the way 2:08:46
that language is used specifically with regards to pronouns and the like
2:08:51
so in this example the doctor ran because he is late the doctor ran because she is late 2:08:56
the nurse Ryan because he is like the nurse friend because she is late the co-reference resolution and the sort of
2:09:02
probabilities of those were were widely varied um because of the sort of statistical 2:09:09
incidents of those types of phrases and this has been shown to bias uh the 2:09:17
assumptions of large language models in and how they detect and how they sort of categorize 2:09:24
different assumptions [Music] there's a few ways of
2:09:30
identifying where the bias comes from there's bias as mentioned the source corpora so if you have input data and
2:09:37
the input data uses gender language the gendered language is interpreted differently 2:09:43
from one another and is is used as part of that language model it can be shown 2:09:49
in labeling so if if people are labeling things differently and they answer a question with one gender differently
2:09:57
than the question of another gender as an example that can introduce some bias 2:10:02
the selection of human Raiders and so if we're selecting human Raiders from a certain from certain population from a
2:10:08
certain part of the world we can introduce biases that way the biases that are introduced by human raters and
2:10:14
test data sets which may be more heavily critical of the algorithm and in terms 2:10:20
of especially identifying the efficacy of particular algorithms can be influenced by the use of gendered
2:10:27
language especially when resolving assumptions and then bias can be introduced by the scientists and model
2:10:33
tuning by doing things like judging the model based on preconceptions of norms 2:10:40
that are used to tune hyper parameters so what do we do with that the way we 2:10:47
mitigate that is through doing things like replacing gendered language and linguistic references to race and
2:10:53
ethnicity with a neutral equivalent or multi-directional expansion for example for any gender language we just turn he
2:11:00
or she into they there's also additional methods through something called bias regularization which uh minimizes the
2:11:07
projection of neutral words on gender and racial axes so essentially as gender and race become identified axes within
2:11:15
the the embedding model those good essentially de-emphasized specifically 2:11:20
with sort of correlation neutral words in occupational job specific again we 2:11:25
just sort of replace occupational job task language with neutral equivalence and then the other sort of interesting
2:11:31
thing is through bias testing and this is my understanding is that this has been shown to improve to actually
2:11:37
improve the model so by removing bias we actually get a more specific Model A more effective model where we introduce
2:11:43
a bunch of bias based testing and try to minimize bias specifically as a new set 2:11:50
of parameters during hyper parameter tuning and then it just ensuring that our input data is diverse that our
2:11:57
Raiders our human raters are are diverse and that we are being representative in
2:12:04
in our data sets with regards to explainability we also
2:12:09
have additional ways to describe the decisions that are being made in a little bit about why they are being made
2:12:15
and so we use generative AI in conjunction with classifiers to to
2:12:21
interpret various alignment scores that are being made and then to disclose what those alignments are and the causes of
2:12:28
those alignments and this allows us to explain the dimensions upon which two 2:12:34
things are matching with one another which assists Us in explainability of alignments 2:12:40
so those are all the things that are going on behind the scenes when we upload the training jacket in terms of addressing some of the problems that can
2:12:47
occur what is happening as well behind the scenes is that we have models that are taking that training jacket are
2:12:54
classifying all the the pieces in it and what we get out is a set of items that 2:13:01
are in sort of Naval language of what the person has done so courses everyone 2:13:08
knows what this are this is someone completing a course but the military especially the Navy uses a lot of of
2:13:14
abbreviations and a lot of coded language in order to explain things and so Osa school is one of those courses
2:13:20
and there are topics that are covered in that course that map over to essentially basic ship operations this is these are
2:13:27
things like navigation steering and and the like communication radio signals 2:13:33
Etc as well there's a an NEC that stands for a naval enlisted classification and 2:13:39
again you know all of these things have corollaries in the commercial and sort 2:13:44
of traditional education sector this is a navalness of classification is much more like a credential and so that Maps
2:13:51
over to the shipboard management credential from me tags a pqs is a
2:13:58
qualification standard and so that is essentially a call qualification and
2:14:03
that Maps over this this particular what happens to map over to an electronic navigation the pqs itself is in this
2:14:11
case redacted it wasn't entirely sure if I was able to disclose which one would be mapping over to this but we have to
2:14:17
deal with these types of problems as well of sensitivity of various pieces of data and that's a an aspect that we
2:14:24
would be wanting to ensure that we follow the right sort of procedures for as well there are deployments within
2:14:30
that that training jacket and so in this case Susan was deployed on the dv680 2:14:36
ddg86 Shoop which is a destroyer and that Maps over to essentially just being 2:14:42
shipboard that's actually strangely enough a sort of experiential credential that we'll talk about a little bit when
2:14:49
we identify the source of it and it actually Maps over to a license
2:14:55
there's a watch station for a maneuvering board operator this is a maneuvering board or a mill board is a
2:15:01
sort of near shore ship navigation tool for keeping track of where other ships 2:15:07
are with respect to your ship and then changing heading direction making sure that you're navigating clearses and
2:15:14
incorporating a bunch of data that is used to create a sort of picture of of what's happening but it's a manual
2:15:21
method and so that Maps over to near shoreship navigation practices
2:15:26
then there's a Navy values course and that that one happens to translate over to 21st century skills because a lot of
2:15:34
sort of basic information about being a reliable sailor is covered in that
2:15:40
particular course so all of these are essentially extracted from the training jacket and 2:15:48
then articulated over to Frameworks and credentials within the confidence and 2:15:55
skill system and credential engine using approaches that were developed during a NSF project called skillsync where we
2:16:02
developed ai-based alignment tools for articulating between different types of 2:16:07
and sort of different languages of credentials and and courses
2:16:14
and those types of things so there we hit next and
2:16:21
now we understand that hey these are the claims that we want to actually go request so it's it's sort of without
2:16:29
question that the military and the Navy has issued these credentials to Susan 2:16:35
um but she wants to get sort of Civilian equivalence and so instead of saying hey 2:16:40
let's send these claims to the Navy and they can validate or verify these claims we want to in this case articulate them
2:16:47
to public and and other types of claims that are granted by various organizations in this case it's ship
2:16:53
operations basic training from the maritime Institute of Technology ship and ship board management from the same
2:16:59
Voyage finding electronic navigation from the same a sailing license from the 2:17:04
maritime licensing agency near shore ship Navigation course from U.S sailing 2:17:09
and the 21st century skills credential from nocti so in this case what we do is 2:17:14
attach the the sort of evidence from the training jacket to each of these and 2:17:20
send the sort of verification requests over to these organizations to say Hey you know we want to try and articulate
2:17:27
these over over time as the articulation Agreements are are sort of put into 2:17:32
place and the Navy has a bunch of these already you know more and more courses credentials and and whatnot can transfer
2:17:40
and so the Navy already has something called Navy cool which is credentialing opportunities online that does a lot of
2:17:48
this type of stuff but it does it from a sort of perspective format in terms of like if I'm a sailor who does this what
2:17:55
can I do later and sort of job patching and things like that this is getting into the actual articulation and the
2:18:01
operational execution of that articulation into these verification requests
2:18:08
so we send those off and now we jump out and jump back into the system as the 2:18:14
maritime Institute of Technology and here we are verifying the claims that they received and so we see hey ship
2:18:21
operation the the ship board management and Voyage planning and electronic navigation we have the sort of evidence
2:18:27
associated with that and we're requesting some some confirmation at
2:18:32
this point you know the maritime Institute of Technology could probably contact Susan contact the Navy if they
2:18:39
need to In terms and do what's necessary to validate those credentials before 2:18:44
they they issue them and so in this case all of those went through everything is 2:18:50
okay confirmed confirmed confirm and we verify those claims coming back in as Susan Malik with our
2:18:56
verified claims we get the sort of fully defined claims here it's Maritime
2:19:03
Institute Technologies claiming that Susan has completed an equivalent of ship operations stcw basic training and
2:19:10
as part of that claim there's you know the associated evidence that comes from the training jacket as well as any other
2:19:16
sort of any other digital artifacts that that we can attach on to that and all of the associated cryptographic signatures
2:19:23
for validation of that claim that anyone may need to do so at this point Susan has the 2:19:29
opportunity to add each of these claims to their ler and we can move forward 2:19:35
so finally we're going to take a look at the ler itself and we can see that Susan 2:19:42
who is currently an os3 has all of these sort of credentials that have been 2:19:49
articulated from her originating credentials within the Navy and this becomes a sort of a learning 2:19:56
and employment record that can get sent to any prospective employer or anyone you know maybe a credentialing
2:20:03
organization or a college actually wishes to attend and can be used to
2:20:09
further her civilian career
2:20:20
left and um thank you for your time [Applause]
2:20:34
all right uh so we're gonna have a a little bit of a break uh we'll reconvene and get started about 3 30. uh just a
2:20:41
couple reminders during the break one uh the QR code on the left the the black one uh is a form for you to fill out if
2:20:49
you want to provide feedback to any of the project teams in any capacity and or 2:20:55
ask a question for our panelists to discuss uh at the end of this so we'll
2:21:00
reconvene it about 3 30 so enjoy the conversations
2:21:20
weeping through the glass counting down the hours
2:21:29
before I hit the bottom
2:21:36
steps in the sets
2:21:48
again
2:22:03
sunrise
2:22:23
[Music]
2:22:29
[Applause] [Music] [Applause]
2:22:35
[Music] [Applause] give us another chance
2:22:59
foreign
2:23:05
[Music]
2:23:33
sunrise [Music]
2:23:59
[Applause] [Music] I know I was wrong
2:24:14
another chance
2:24:22
[Music]
2:24:56
[Applause] oh Mr pit oh Mr pit Mr Pitiful who let
2:25:03
you down who let you down who let you down
2:25:11
you still don't believe you don't believe you don't believe in your 2:25:16
grievances show unfolds
2:25:23
[Music] Cloud you see
2:25:29
the city oh you admit that you were wrong [Music]
2:25:38
it don't make me feel bad that we're still friends [Music] 2:26:02
silence my friends
2:26:11
so where did you go where did you go where to go
2:26:17
[Music]
2:26:25
well I don't believe I don't believe I don't believe everything I see 2:26:33
and if you don't like the movie [Music]
2:26:41
Cloud you're sitting on I don't expect you to admit that you were wrong 2:26:50
I just wanna know how you don't make me feel bad that 2:26:56
we're still friends [Music]
2:27:13
to make amounts still be
2:27:19
[Music] friends
2:27:26
still be my friends still be my friend
2:27:35
still be my friend [Music]
2:27:45
ha ha ha ha ha ha ha ha
2:27:51
foreign [Music]
2:28:19
[Music] said we are not we are not shining stars this I know
2:28:28
I never said we are though I never been through hell like that of close enough Windows to know you
2:28:35
can never look back [Music] if you lost in alone or you're sinking
2:28:43
like a stone carry on
2:28:48
may your past be the sound of your feet upon the ground carry on
2:28:57
carry on carry on
2:29:03
[Music]
2:29:21
our neighbors
2:29:26
left [Music]
2:29:40
[Music] I'm not that you are
2:29:46
beautiful [Music]
2:30:06
[Music]
2:30:17
door hold the phones show me
2:30:34
[Music]
2:30:54
[Music]
2:31:01
[Music] [Applause]
2:31:10
[Music]
2:31:21
foreign
2:31:28
[Music] [Applause] foreign
2:31:36
[Music]
2:31:53
[Applause] [Music]
2:32:05
[Applause]
2:32:11
[Music]
2:32:27
oh yes this is powerful stuff and got me circling like the moon around 2:32:35
the Sun acting crazy [Music]
2:32:45
stuff there's no way for you to give it
2:32:50
up [Music]
2:33:02
and keep your turning like it's never enough
2:33:08
all right now let's turn it up
2:33:13
dare do like a flower does [Music]
2:33:19
she opens up as Sunny Rises
2:33:24
[Music] I'll hit me this is
2:33:30
[Music] power powerful
2:33:39
powerful [Music]
2:34:02
foreign [Music]
2:34:19
[Music]
2:35:25
[Music]
2:35:41
thank you okay
2:35:50
all right can we get people to start making their way back to their tables
2:35:57
we're gonna move on to our next use case
2:36:04
so if you could start making your way back to your tables that would be wonderful we're going to transition to our fourth
2:36:11
and final use case uh introduce a Persona and hand it to our next group so if you can make your
2:36:17
way back to your table that would be wonderful thank you thank you
2:36:23
[Applause] how do you think that's going to go how long how long
2:36:40
all right so I'm going to start with the next Persona slide as you quietly make 2:36:46
your way back to your table we've got two more presentations from our technical teams 2:36:52
and then we will have about a 45 minute panel conversation with five panelists 2:36:58
and uh the room you so if you have not already gotten your questions ready or 2:37:04
taken some notes on a few of the you know most impactful parts of the
2:37:10
presentations uh now's your chance to start doing that and those notepads on the table and the pens and everything
2:37:16
I believe you're yours I don't work for the chamber but they put them out there so I assume you can take them
2:37:23
all right our final use case for today's
2:37:28
presentation is built around the experienced but unemployed worker
2:37:34
one of our project teams which is coming up next
2:37:40
was very focused on the incarcerated person this is a very unique population and 2:37:47
some of the the limitations around what this group was able to do inside of Prisons with this technology was very
2:37:55
very constrained it is this use case that I think really
2:38:00
highlights some of the challenges but also some of the importance of really
2:38:06
rooting your development in people
2:38:11
working with humans who are going to be using this technology in these solutions to make sure that you can understand to
2:38:18
the best of your ability their situation their experiences the challenges but also the opportunities that arise when
2:38:24
you're able to bring their skills to the surface our next group is the work bit is work 2:38:30
Bay and they worked in collaboration with participate in California State University Dominguez Hills
2:38:37
work bays and interactive Career Development training and recruitment platform that links job posting and
2:38:43
applicant tracking skill building and career navigation and one system delivered via web and mobile IOS and
2:38:50
Android applications workbay provides a localized customized solution of support for economic 2:38:56
stakeholders to achieve their goals participate is a platform and service
2:39:02
provider that creates Solutions specifically designed to support and operationalize communities of practice
2:39:09
Cal State University Dominguez Hills Workforce integration Network uses advanced technology technological
2:39:15
resources combined with University and Industry generated business data to demonstrate the benefits of the
2:39:21
University's educational experiences before the benefit of all stakeholders the project lead on this is Mary Hayes
2:39:28
and I know she's not coming up on stage but she yeah she definitely deserves a round of applause
2:39:34
as I was describing some of the project leaders and and the Champions Behind these efforts this team would not be up
2:39:40
here without Mary she is a very passionate individual she's an entrepreneur has founded 2:39:45
multiple educational and learning companies with over 30 years of experience in Workforce Development educational publishing teaching and
2:39:52
developing multimedia educational content for industry and government
2:39:58
the presenters for this team I have worked with Dr Crystal Rawls in a variety of working groups over the last
2:40:04
few years and it is an honor to meet her in person for the first time today and introduce her to the stage she's an
2:40:10
entrepreneur return educator and author with more than 20 years of progressive management experience and repeated
2:40:16
successes in developing project initiatives directing project plans and achieving performance targets
2:40:22
Crystal owner of La Candace group is a recognized voice in Workforce
2:40:28
Development developmental practices she is featured in growing fairly how to build opportunity and equity in
2:40:34
Workforce Development and consults with women in technology International and Equitable hiring
2:40:39
practices in the technology industry Dr Rawls is committed to developing Partnerships that help organizations
2:40:45
operationalize their diversification initiatives presenting with her is Stephanie Taylor 2:40:51
Thompson and she's formally a re-entry specialist for the Idaho Department of Corrections and currently serves as
2:40:59
director of re-entry transformation for work Bay having experienced incarceration 2:41:04
Stephanie earned full pardons from Idaho and Montana in 2017.
2:41:10
she is in long-term recovery she's a graduate of a Idaho State University with degrees in sociology and
2:41:16
criminology Stephanie is currently enrolled at Northwest Nazarene University to earn a 2:41:22
master's in social work she is a certified peer support coach and specialist a certified Family Support