Latest / Leaders in Lending / Launching Upstart’s AI Certification Course

Transcript

Launching Upstart’s AI Certification Course · Leaders in Lending

• 0:02
  You are listening to Leaders in Lending from Upstart, a podcast dedicated to helping
• 0:07
  consumer lenders grow their programs and improve their product offerings. Each week, here
• 0:13
  decision makers in the finance industry offer insights into the future of the lending industry,
• 0:18
  best practices around digital transformation, and more. Let's get into the show,
• 0:26
  all right, Welcome back Jeff as a guest to the podcast. So
• 0:31
  Jeff and I are two of the instructors and an exciting new AI certification course
• 0:36
  that Upstart is launching. Happy to have you back on Leaders Lending, Jeff,
• 0:41
  Thank you. What's the saying? Every does a long time listener,
• 0:44
  first time caller, I guess, longtime host, first time guest. But
• 0:48
  it's fun to be back in front of the microphone, so to speak.
• 0:51
  Well, great, Well, I'm looking forward to talking about the AI certification.
• 0:54
  I know we've all been spending a lot of time working on it internally
• 0:57
  and exciting to get it out out kind of in the public, so,
• 1:02
  you know, for any of the listening, Like, you know, can
• 1:03
  you give us some background and like why did you decide to make this course,
• 1:07
  Like what was the what was the intent ofment, and how did kind
• 1:10
  of how did it come about as an idea? I think, you know,
• 1:14
  we kept talking to banks and credit unions and various people in the industry.
• 1:19
  You said, look, the AI thing is coming and I don't understand
• 1:21
  it that well. And it seemed to us like there was so much educational
• 1:25
  content, but it tended to either be at like fifty thousand feet or five
• 1:29
  hundred like either for people wanting to build models and you quickly go, I
• 1:32
  don't understand what this means, or there was content that was like very how
• 1:36
  to use chat GPT to write better sales emails, and the idea of wanting
• 1:41
  to help people better understand at a business level, at a functional level,
• 1:45
  how these things work. Not to be a to be a developer, but
• 1:48
  to be a good consumer, a good user, a good buyer, or
• 1:51
  a good overseer of the models felt to us like a gap in the good
• 1:56
  educational content. That's what we were hearing from from partners and customers, and
• 1:59
  we thought, well, I mean, it seems like something we could do,
• 2:00
  and maybe it'd be useful if we just went out there and tried to
• 2:04
  help people better understand this space a little bit. So that was really once
• 2:07
  it was my idea, I think a number of people and I started,
• 2:08
  we're seeing this and said, hey, I wonder if we could do something
• 2:12
  valuable and contribute to the space and help people understand a little bit better.
• 2:15
  Sure, so, you know, I definitely think there's no shortage of at
• 2:20
  least content of various levels of AI out there. But who who's the who
• 2:23
  do you see as the ideal user of this course? Like, what's the
• 2:27
  right persona of people who may really get something out of taking this class from
• 2:32
  the team? Yeah, well so I guess really, really anybody in financial
• 2:37
  services, you know, and really focus on business people. I don't want
• 2:39
  to I don't want to start with who the course isn't for. But we
• 2:42
  really didn't want to design a course that's intendo like how to build a model,
• 2:46
  how to pick the right model form. We wanted to say for those
• 2:49
  of you who are maybe looking at an internal team building a model and trying
• 2:52
  to understand can this be used in a process I run or in a program
• 2:55
  I manage. Anybody in risk or compliance or that runs a line of business
• 3:00
  within a financial institution or oversees some functions and wants to be able to ask
• 3:05
  better questions understand the model. That's really what it was for. So really
• 3:07
  targeting that community of bankers, of credit union executives, anybody in the kind
• 3:13
  of business side of that who wants to have a functional understanding of AI,
• 3:16
  that's really who we built it for. Not that it's limited to those people,
• 3:20
  but if you think about who is the audience, it's people who are
• 3:22
  interfacing with AI models, with teams building AI models, but aren't people actually
• 3:27
  building models themselves. And that was who we wanted to really empower and educate
• 3:30
  through the course. That was our kind of focus in terms of the audience.
• 3:34
  So really to help kind of the non technical maybe business owners have a
• 3:38
  better grasp of the technical background of what is happening with the AI models and
• 3:44
  thinking about how they could use them to improve or enhance their business. Absolutely,
• 3:47
  And if you think about I mean, I hear some people in compliance
• 3:51
  that go these teams brought me an AI model, and like, I don't
• 3:53
  know what they're They're using these words to describe what it is, and they're
• 3:55
  not words that I'm familiar with, or they're described things in different ways.
• 4:00
  How do I get a better grasp on what questions should I be asking or
• 4:03
  how do I evaluate if this is an appropriate model for this use that kind
• 4:08
  of thing. We thought, that's that's the audience we want to empower.
• 4:10
  There's a lot of content if you're building models that's out there, and if
• 4:13
  you're kind of a consumer level user of models, I think there's a lot
• 4:15
  of content out there as well. But at that kind of buyer, overseer,
• 4:19
  you know, manager of a function that's using an AI model in some
• 4:23
  way or evaluating one, that's really the audience we wanted to reach out and
• 4:27
  give some better structured content to sure, so kind of the decision maker really
• 4:31
  at the business who's going to be kind of owning the business outputs of these
• 4:36
  things. So, you know, can you give a little bit of an
• 4:39
  overview of the kind of maybe the modules or kind of key learning objectives that
• 4:43
  are covered in the certification course. Yeah, we can go through the modules,
• 4:47
  but if I think at a higher level about what we wanted people to
• 4:49
  walk away with, it's really an understanding of how these models work, how
• 4:55
  they're built, how they function internally, how to assess the suitability of a
• 5:00
  model for a given purpose? Right, So, not so much so I
• 5:03
  use an XG boost model or a neural in that model. But hey,
• 5:06
  given this model, like does it have the right kinds of inputs, does
• 5:09
  it have the right kinds of outputs of the accuracy metrics on par with what
• 5:13
  I might need in this context? And how do I think about how to
• 5:16
  use that model in my world? What kind of guardrails might I might put
• 5:18
  on? So really starting with how how do the models work? How do
• 5:23
  I assess the appropriateness of a model? How do we manage and oversee there's
• 5:27
  a whole section with any our chief compliance officer talking about governance and oversight and
• 5:31
  model risk management and all those regulatory concerns, And then there's there's a little
• 5:35
  section at the end where we kind of get into, like i'll calm brain
• 5:39
  ticklers on how is THEI being used in financial services? What are ideas for
• 5:43
  where this might be applied? And also what can I be doing to help
• 5:46
  my organization get ready for an aipower future? How do I build the foundation?
• 5:51
  So kind of running the spectrum from a little bit. There's a little
• 5:54
  bit of technical content in there, and so we wanted that to be available,
• 5:58
  But then there's a lot that builds on that in terms of how you
• 6:00
  manage or oversee or evaluate and assess a model, and then how's it can
• 6:03
  impact financial services? And where do I think about getting started and building an
• 6:08
  organizational capacity to leverage these technologies moving forward? Hey there, former host Jeff
• 6:14
  Kelner here to let you know about an exciting opportunity to strengthen your understanding of
• 6:18
  artificial intelligence and financial services. If you want to better understand the fundamentals of
• 6:23
  how AI models work and how they will impact the financial services industry, your
• 6:28
  in luck. Upstart is launched an online course and certification designed to do just
• 6:31
  that. This course will give you the knowledge you need to lead the conversation
• 6:35
  about how your organization can best leverage and manage this revolutionary technology. If that
• 6:41
  sounds interesting, check out the course as it's currently free for a limited time.
• 6:45
  You can learn more at upstart dot com slash AI Certification. Again,
• 6:48
  that's upstart dot com slash AI Certification. Thanks, and now back to the
• 6:54
  show. I definitely think that a lot of the content really kind of demissed
• 7:00
  find some of the areas. So I think there's a lot of great explanation
• 7:03
  of various areas of creating a model using it and as you said, real
• 7:08
  like practical tools. What can you do with this? How can you use
• 7:11
  these these AI models to potentially drive your business forward or create new opportunities?
• 7:16
  So I know where we were doing the recording and putting the material together.
• 7:24
  We had a studio set up in our office and had a team in uh
• 7:29
  recording each of us kind of talking through our various topics. What was your
• 7:35
  experience as an instructor in the course, Like, what did you uh uh,
• 7:40
  what did you find interesting about it? Well, I answer, but
• 7:43
  I want to hear your answer because you're you're instructor in this course too.
• 7:46
  I mean I got to see the whole thing, soup tons, which is
• 7:48
  I was really involved in structuring a lot of the content, what are the
• 7:51
  things we want to cover, what should the modules look like? And then
• 7:55
  writing a lot of the words. And it was always surprising to me.
• 7:58
  I think it's too funny to if it's one like reading your own words and
• 8:01
  hearing them out loud. The number of times I got in there and went,
• 8:03
  oh my god, that doesn't sound the way I thought, or the
• 8:07
  producers would go, can you give us another take? Would sound more podcasting
• 8:09
  and less professory and I got, I don't know what that means. I
• 8:13
  don't really know how to voice act in that way. So that was a
• 8:18
  really kind of odd experience to be in there. And then the second really
• 8:22
  funny one, and I felt bad for some of our our Upstart team members
• 8:24
  because you know, there was hair and makeup. You guys got hair and
• 8:26
  makeup for the ladies. But the guys they just brought in a little bit
• 8:30
  of foundation for us and there was no one to apply it. And the
• 8:33
  ladies that were looked at me and go, do you know how to put
• 8:35
  makeup on? I said, I said no. So these poor these poor
• 8:41
  people were like a coming in putting foundation on my face. And they weren't
• 8:43
  producers Upstart. You know, Ana was there, and so they were trying
• 8:46
  to get my face to not shine so much, which I guess I'll have
• 8:48
  a problem with. So that was kind of a funny moment. But I'm
• 8:52
  curious what your experience was because you more got content that had been written,
• 8:56
  so you were both reading to some extent someone else's words, although I'm sure
• 8:58
  slightly edited, and then you had your own experience in studio. Definitely,
• 9:03
  I think the you know, it was interesting reading the scripts, and I
• 9:07
  think one of the things that I learned the most is what sounds and sounds
• 9:11
  logical and makes sense on paper. Things like bullet points are numbered items.
• 9:16
  You have to say that a lot differently as you're talking about it. You
• 9:20
  can't just read through it or referring back to the previous section. That doesn't
• 9:24
  work in audio, And so going through that and you know, rewording things.
• 9:30
  And then as you said, the producers put more inflection on this,
• 9:33
  have a little bit more excitement here. And I told them, I'm like,
• 9:37
  well, you know, we're all kind of probably relatively type A in
• 9:41
  some ways, we can only be so exciting. You know, we're not
• 9:43
  actors, and the material is pretty pretty serious. But I think my probably
• 9:50
  my favorite one that I think if you take the course, you'll see me
• 9:54
  probably laugh a little bit as I'm reading it, was it was obviously a
• 9:58
  clear Jeff wrote this section of the script. There's a whole section on a
• 10:03
  cal Tech and a fish study done and and the different i think patterns of
• 10:07
  fish migration. As I'm reading this, I'm thinking, I definitely didn't write
• 10:11
  this, but I know who wrote this. Uh, Materials issue of the
• 10:16
  cal Tech that gave me away that I have. Yeah, it was a
• 10:18
  little bit of both. And so I think when I watched the video back,
• 10:20
  I see myself kind of laugh as I'm talking about fish and cal Tech
• 10:24
  and like, this is great story. But it was a great analogy,
• 10:26
  but definitely not one that that came to my mind. And you know,
• 10:31
  I do think Anna was very excited and very happy to be there and produce,
• 10:37
  so it was it was. It was a fun time as well as
• 10:39
  is just getting really a lot of really great content out there that I think
• 10:43
  will be really, really valuable. I was surprised how different. I mean,
• 10:46
  I've done a lot of presentations in my day, but usually a lot
• 10:50
  and so there's there's no retakes, there's no hey, you stumbled over these
• 10:54
  words? Have you slurred a little bit? Here? Let's go. And
• 10:56
  I was a little i won't say taken aback, but I was a little
• 11:00
  surprise at how often the priestser goes that was that was all right, but
• 11:03
  let's uh, let's try that again. And I'm just not I'm used to
• 11:05
  having to deliver once and just live with the results and being confident with that
• 11:09
  delivery, and then having the once they stop me a few times to go
• 11:13
  back and try something again. Then it got in my head like, man,
• 11:15
  you gotta stop doing these stupid things, and then it would happen more
• 11:18
  and they go, well, you know, I think you can do better,
• 11:20
  said okay, okay. That was an interesting It's just a very different
• 11:22
  experience than being on a live call or doing a webinar or speaking at a
• 11:26
  conference, where you it just it's going to be your delivery, is your
• 11:30
  delivery, and then it is what it is, and here they're like,
• 11:31
  you know, we can take it again and do it better with a little
• 11:33
  different voice exactly now. And I think we spent a lot of time sitting
• 11:39
  in front of the camera and working through it. And I think during during
• 11:43
  a number of my sessions, there were we think it was either planes or
• 11:46
  construction that kept kept making noise, and so I'd be talking and then they'd
• 11:50
  say hold on and they'd put the hand up and we'd pause, and we'd
• 11:54
  wait about ten minutes until until something came else came out and h and then
• 11:58
  reached art. So definitely takes a lot longer to get through through something like
• 12:05
  that than a just kind of a live live presentation. But I do think.
• 12:09
  I think the content's great, very excited for it. I think,
• 12:13
  are what are you most hoping that students who take the course, like,
• 12:18
  what would you like to see people take away from it that where they really
• 12:22
  feel it adds value? Yeah, it's I really want people who take the
• 12:26
  course to come away feeling empowered to not just engage intelligently in the conversation about
• 12:33
  how and where to use AI and what the risks are and what the upsides
• 12:35
  are, but to really feel empowered to lead those conversations. I remember when
• 12:39
  I first got into AI stuff. I'd be sitting in the room going,
• 12:43
  I don't don't understand what they're saying, and I didn't feel able to contribute
• 12:46
  or question. Or I might but that didn't make sense to me. But
• 12:48
  I feel like I don't have enough background to question what is being said about
• 12:52
  this, Or they say this risk metrics okay, and I go, I
• 12:54
  don't see why. But I also didn't feel comfortable saying saying that in the
• 12:58
  room, And so my real goal is that people walk away going, hey,
• 13:01
  I feel informed enough on these topics. When the executive team is discussing
• 13:05
  this, or we're debating if this fraud model is working effectively or not,
• 13:09
  or what the risks might be that I'm informed enough to go and have an
• 13:13
  opinion, to ask a smart question, to call in to question something maybe
• 13:16
  a technical team is saying. And really, I mean it feels to me
• 13:20
  like an inflection moment in the usage of AI. And so I want people
• 13:24
  to feel empowered to help lead the charge and how and where to use AI
• 13:26
  effectively in financial services. And I hope that's what they come away from this
• 13:30
  course with absolutely so, and I'm glad we can be part of empowering hopefully
• 13:35
  a large audience of people to feel that and to learn more about AI and
• 13:39
  how it can benefit them. So, you know, for anyone who's listening,
• 13:43
  where can they go if they want to find more resources and then enroll
• 13:48
  in the course, where can they go to get that information? Yeah,
• 13:50
  So the course just give the real basics. The course is online video based.
• 13:54
  It's a couple of hours long, I think four or five total hours
• 13:56
  of content, but it's broken up into pretty short videos on the order of
• 14:00
  I have the ten minutes that are very topical, so our goal was digestible,
• 14:03
  and then there's a PDF takeaway with a lot of the key points that
• 14:07
  I really push the team to think of as a job aid, something that
• 14:11
  you would have and go, hey, I learned these things. But now
• 14:15
  I'm in the moment, I'm in the meeting. I'm prepping for this meeting
• 14:16
  about an AI model, Like what was that framework for evaluating appropriateness or for
• 14:22
  what were those accuracy metrics? And how to think about it. I want
• 14:24
  to have that on my desk to be ready to really engage in the conversation.
• 14:28
  And so a couple of short you know, got the videos that are
• 14:31
  digestible, link the pdf, and then there's an online community that people have
• 14:35
  as well. Will be a LinkedIn group that you can engage that. We
• 14:37
  will have instructors available to ask questions of if you want to dive deeper or
• 14:41
  to understand something. And then our goal is really community of practice that people
• 14:45
  who take the course and are doing this in the field and running it.
• 14:48
  Hey, here's how we talk to our compliance officer about this. Maybe that'll
• 14:50
  work for you, or hey, i'm struggling with getting a certain stakeholder on
• 14:56
  board. Has anybody else seen that? And so we're hoping to really support
• 14:58
  that communit. You can find out about the course at upstart dot com slash
• 15:03
  AI certification so you can see the information register for the course. It's free
• 15:07
  for a limited time, so I recommend if people are interested or find this
• 15:11
  topic interesting or useful for themselves, it's a great time to check it out
• 15:16
  because it won't be free forever, but it is for now. Great Well,
• 15:18
  thanks Jeff for joining us today. I think we're definitely excited for all
• 15:24
  of you listening to register and to learn more about AI alongside our instructors.
• 15:30
  Absolutely, it's been a lot of fun making it, but I think the
• 15:33
  real fun starts once it gets out into the world and we get to start
• 15:35
  hearing from actual learners and students and interacting with the community. So we're very
• 15:39
  much looking forward to phase two because Phase one has been a lot of work
• 15:41
  and a lot of fun, but I think Phase two is going to be
• 15:43
  a lot more fun. So we're looking forward to talking. Everybody wants a
• 15:46
  good chance to see all the content. Upstart partners with banks and credit unions
• 15:50
  to grow households and expand consumer lending. Through its leading AI lending platform.
• 15:56
  Upstart powered banks and credit unions leverage AI to offer higher prooval rates and experience
• 16:00
  lower loss rates while simultaneously delivering the exceptional digital first lending experience that consumers demand.
• 16:07
  Whether you're looking to grow and enhance your existing personal and auto lending programs,
• 16:11
  or you're just getting started, upstart can help. Upstart offers an end
• 16:15
  to end solution that can help you find more credit worthy borrowers within your risk
• 16:21
  profile. With all digital underwriting, verification, loan closing, and servicing,
• 16:26
  It's all possible with Upstart in your corner. Learn more about finding new borrowers,
• 16:30
  enhancing your credit decisioning process, and growing your business by visiting upstart dot
• 16:34
  com slash Lenders. That's upstart dot com slash Lenders