Roads Taken

Model Intelligence: Will Uppington on building skills and looking inside the models

Episode Summary

After a pivot away from medicine to the business world, Will Uppington still needed models to show him what that world was all about. Through a stint in consulting and a series of technology start-up experiences, he gained the knowledge he needed to start his own venture. Find out just how important models really can be and how knowing what is inside them is what gives them power.

Episode Notes

Guest Will Uppington, Dartmouth '96, had a desire to do good in the world, so naturally thought he would follow in his father’s footsteps into medicine. But after a few too many hospital experiences he thought there might be another way to be of service and pursued a double major in economics and government. Worried this might not give him quite enough hard skills, he spent senior year dabbling in Chinese and math and coding. Despite these new skills and yet without any role models in the business world, he still sought a surer footing and so began is career in management consulting.

Consulting experiences in the technology sector and newly-deregulated energy industry led him to a stint in venture capital, just as the dot-com bubble was bursting. Even so, he knew he wanted to pursue a life in technology entrepreneurship after business school. A few start-up experiences later, he was ready to launch his own venture, this time looking more deeply at the models in front of him, namely in artificial intelligence and machine learning.

In this episode, find out from Will how important models can be and how knowing what is inside them is what gives them power on ROADS TAKEN...with Leslie Jennings Rowley.

 

About This Episode's Guest

Will Uppington is CEO and co-founder of Truera, which helps enterprises analyze machine learning, improve model quality and build trust. Find out more at truera.com. Though his first founder role, this certainly isn't his first rodeo in the technology start-up world. He has a started and grown a number of successful companies and products and always tries to do something that is both of personal interest to him and of use to the greater world. He lives in the San Francisco Bay Area with his wife Lauren Currie Uppington ’96 and their family.

 

EPISODE TRANSCRIPT

 

Executive Producer/Host: Leslie Jennings Rowley

Music: Brian Burrows

 

Find more episodes at https://roadstakenshow.com

Email the show at RoadsTakenShow@gmail.com

Episode Transcription

Will Uppington: The arc of history right now is that absolutely know technology advancements are always going to be critically important for improving the standard of living for people and solving a lot of society's problems. And so I just felt it was the long term right thing. I was proud of myself for what I was interested in. And that I was passionate about. And also something that can contribute to the world and the markets will come around.

Leslie Jennings Rowley: After a pivot away from medicine to the business world, Will Uppington still needed models to show him what that world was all about. Through a stint in consulting and a series of technology start-up experiences, he gained the knowledge he needed to start his own venture. Find out just how important models really can be and how knowing what is inside them is what gives them power...on today's ROADS TAKEN with me, Leslie Jennings Rowley.

Today, I'm here with Will Uppington and he's going to talk about making his way in the world in a slightly different way than he had anticipated...like all of us. And it's just lovely to have you here, Will

WU: Happy to be here. Thank you. 

LJR: Great So we're going to start the same way, I start all of these the same. Will, when we were in college who were you. And when we were getting ready to leave, who did you think you’d become. 

WU: So in college, I spent a good amount of time in both academics and athletics. So I was on the cross-country indoor track and outdoor track teams. So it was a three season athletic endeavor. And that took a good amount of time with practices and meats and things like that. And then I first started coming into Dartmouth thinking I was going to go into medicine. My father is a doctor. And I was had worked in hospitals during high school and felt that medicine was my calling. So I started out being premed. And then after working a little bit too much in hospitals where I did not have the most glamorous job, it involved cleaning equipment and things like that. After operations, I decided maybe medicine wasn't my calling. And so I then just pursued what I was interested at the time, which was economics and government. And that's what I decided to do, my double major in economics and government. So that was essentially most of my time at Dartmouth. It was mostly split between academics and athletics. There was a few bits of time for some fun as well. That would be very much in the typical nature of most students. So I probably don't need to go into too much detail.

LJR: No, We know that. And well, some of the fun actually led to your marriage, right?

WU: That's right. I'm married to another fellow Dartmouth 1996 graduate. Lauren Currie was her name back in college. And we met we actually did a foreign study abroad program in London and definitely benefited. I love those programs. I think that's a really nice aspect of Dartmouth. And then that's where we first met. And then we started dating our senior year. And then and then kept going and was and part of that first flight of marriages that happened after undergrad if you find your person in undergrad. That's right. And so, yeah, that Dartmouth has been obviously hugely impactful in my life and in more ways than one. 

LJR: Yeah, well, and I think that unlike many of us who did not have that great realization of finding the one, then it really did help shape where you were going and what you were doing. So as you were leaving, where did you think those next few steps would take both of you? 

WU: Actually, Yeah. So as far as we were leaving, it was interesting because I had, I would say, a third transformation within Dartmouth, because after I decided to double major in government and economics, I came to this realization that I might be leaving Dartmouth without some hard skills. So actually, my senior year, I decided to take Chinese and computer science courses and some operations research and math courses. So it actually was quite the opposite of what other people did. My senior was not the most fun going to Language Lab early and then doing a bunch of introductory math and computer science courses. But I did feel at that point that my path would take me more into technology over time. And that's why I did some of those courses, but I had not really taken enough to fully make that transition. So as I was leaving Dartmouth, I selected consulting as kind of a way to go learn about the business world since I had not had a lot of mentors in the business world coming from a family of mostly doctors and professionals like that, but do it in a somewhat technical way. So my first job out after Dartmouth was doing consulting, but we were doing, I was running linear optimization models to predict potential electricity market prices during that time of the electric industry was going through some deregulation and there was a lot of need for help in terms of what's the new market price is going to be for electricity, which had never been market driven before. And then what are the implications of deregulation? Where should we build power plants and things like that. And so I did consulting for that company, which has now since been acquired, and then at Boston Consulting Group did a bunch of consulting in that area. 

LJR: It seemed like the right thing to do. And you're still kind of in that mindset of skill building. I'm sure you learned a ton from that, but the kind of the stair step progression out of those early years of consulting is often business school. And that's your path, too, right?

WU: That's right. So I actually, at Boston Consulting group, I did a bunch of work on energy, but I also started doing some work in technology because I was out in the Bay Area. And there were a lot of technology clients out in the Bay Area. And that's what I became interested in. I did make a decision at that point, which was do I want my career to be in energy related field or do I want my career to be an information technology related field, and I felt that while my heart was very much in energy and still is in terms of that being a great way to contribute to the world, I felt that there was so much regulation there that it would be just really slow to make progress. Whereas in information technology, there's much less regulation. And you can do things much the faster, you can accomplish things much faster. So I then went into venture capital actually after that, where I was doing both information technology and also energy and investing. And then after did a couple of years of venture capital, which was a very interesting time because I joined venture capital in 1999 and at that time, there was a very big bubble going on – the famous internet bubble – and then that bubble burst and then I was doing venture capital in the aftermath of the bubble bursting. 

LJR: That's educational. 

WU: Yes, it's a great, great education to see both the highs and the lows, and the kind of lemming mentality that can occur during bubbles and then see the aftermath of that. And the almost the pendulum swing to the opposite end of the spectrum, where things that did make sense weren't getting funded just because people were scared in the aftermath of that bubble bursting. So I went, I did at that point in 2000 to go to business school. I went to Harvard Business School with my wife, Lauren, and I used that, I made essentially the decision to go into software entrepreneurship at that point in time. And I used business school as a way to facilitate that transition to where I was going to go from being an essentially consulting in venture capital, helping people to do other things. And then I wanted to go do those things myself, go build companies myself and I, and that's how I used business school as a way to enable me to do that. Plus, business school is a great life experience. And I figured the worst came to worst is that even if I kind of retire two years later, because I spent two years in business school, what would I want to do with that time? I'd love to do something like business school because you meet a great set of people and you have a kind of a great, great life experience. So we both went as well. That was also part of it. And it was it was a fantastic experience.

LJR: Yeah though I will say in hindsight now, it was totally the right path to be thinking software entrepreneurship. I mean, that seems very logical. But at that moment when you'd seen that bust and you'd seen it from the side of … and now you'd seen, as you said, the shift to the other side of whoa, we're really going to get down to fundamentals and people have to really prove themselves to get backing, like that seems like not the logical, oh, I'll go into software entrepreneurship. What was that thinking to get you to be so kind of forward thinking when you're mired in that like ugh..? 

WU: I know that's a great point. And in fact, you can often track bubbles a little bit by what MBAs are doing and how much MBAs are going into a particular field. And there were a ton of MBAs graduating in 2000, 1999 that were going into technology. And then after 2000 a lot less did. And when I went to business school, that was it was still at a relatively low point for business school graduates to go into technology. So it was a little bit of, like you said, not following the stream, if you will, a little bit and going into a little tributary or something like that. So I think ultimately, my perspective on what people should do is that they should be doing something that is personally passionate for them and that they feel is their way to both enjoy life, but also contribute to the world. And I just felt that there's so much still to be done in information technology or even energy technologies that there's no way, while you may have little bubbles here and there, the arc of history right now is that absolutely, you know, technology advancements are always going to be critically important for improving standard of living for people and solving a lot of society's problems. And so I just felt it was the long term right thing, both from myself and what I was interested in. And what I was passionate about, and also something that can contribute to the world and the markets would come around. 

LJR: Right and that has certainly borne out. So good job being forward thinking. So tell us about where you found yourself. What did you join? And then ultimately what have you created?

WU: Yeah, so after business school, I decided to go into startups and learn how startups work and how you actually build startups. So I joined a company with about 80 people that I thought would be successful because just because I wanted to have that experience of seeing a successful startup, it was both a good or a bad thing. But it turned out I was right in that was going to be successful. But it didn't quite give me the runway because actually, while I was in business school, the company got acquired. So I got my job offer. I agreed to join the company. And then the company got acquired. And then after right after I joined the company that acquired my startup actually got acquired again. So I ended up, after a few months, working at quite a large company called Juniper Networks. I got my job offer from a startup called Notaris, which had built a security and a virtual private networking technology, a kind of a next-generation VPN technology. And then that company got bought by another company called net screen, and then that screen got Juniper Networks. Now, the good thing was, is they kept us as kind of our own team within the big company for at least a year. So I did get some of that startup experience, but I also ended up getting some experience working at a large company. I did that for about 2 1/2 years. And then I joined another startup, which was in analytics and in search technology, and I joined that quite early in around a number of employees, was around 30 and then works there for another three years. The company eventually grew to several people and over $100 million in revenue and ultimately got acquired by another large company. So that was a great experience because we started really from the beginning when there was a there was a barely a V1 product or first version of the product. And I was leading product management. And so really help learn how do you develop a product that is going to be successful when you're starting from essentially scratch at ground zero? So that was a great experience. And then after that, I just went to a next smaller company. So I joined a company where I was the fifth employee. I was the first person outside of the engineering and the founding team. And I teamed up with a great founder and got really the experience of kind of building it with him and the other co-founder. The three of us met every Sunday for four years to make all the major decisions of the company. I built out lots of different teams, built out all the different products. And so that was just a kind of a great, great experience. And then possible to. Now, I was with that company for about nine, 10 years, and then I got that company to a certain point, which I felt could be proud of. And then I'd identified a problem that I thought was worth solving and that led me to actually be a co-founder in the company that I'm at now. 

LJR: So that's really interesting, Will, that you have these kind of shorter stints at startups, got them through their infancy stage and adolescence to being acquired stage. And that seems to be, did that a couple of times, and then you stayed for a really long time in this world, right in one place. And then you're able to make the move. What do you think it was about? Did you did you think you needed that time to see all kind of the full progression of a business staying its own entity? For you to say, I can…not only do I see this problem, but now I know that I can do all of these things? Is that what it was or was it something else that just it was the right time for you to go?

WU: So the reason I say to the third company for a really long time is because I felt an obligation, because I was essentially kind of like a quasi co-founder of the company. And the company actually had and this often happens with companies. The company had quite a journey. So our first product was super successful. Initially, we were one of the fastest growing software companies ever. We went I was looking at the numbers. We went from 0 to about $20 million in three years in revenue, and that's pretty fast. But what we discovered was that the company could sell really well initially. But once you made that initial sale, it was quite hard to sustain the revenues. So we had to do a bit of a pivot. We had to create new products that we thought would be more sustainable. And I really took on that challenge in a significant way. And really architected a lot of the strategy for that pivot. And I felt an obligation and a passion to see that through. You know, when you take on investment from other people, I think that that's quite a significant obligation to kind of do your very best to make that investment valuable for our shareholders. And so we did it. We did amazingly well, but we had to we had to execute this pivot. That's the main reason why I stayed for a really long time, because I was in a leadership role. And I was driving a lot of that change. And fortunately, we were able to. It was very difficult. It was very hard, but we were able to execute that pivot. And the great thing now is that we got the company to quite a good amount of revenue, greater than $50 million in revenue before I left. And then the company just did a recent financing, it just acquired another company, so its financing was done at quite a good valuation, I believe, close to a billion. And so now it's really grown considerably. And that pivot has proved to be successful, though it took a lot of effort and time. 

LJR: Yeah so now you're putting that effort and time into your own baby. So tell us about the new company.

WU: So the company is called Truera. The goal of the company is essentially to remove the black box from machine learning so that people can build better machine learning applications, that they perform at a high quality level. And that we can help them maintain that quality and performance of those applications. So the reason I got involved in this is that as part of my role in leading the product organization at Bloomreach, my last company, we started to use machine learning technology. Machine learning is probably a technology many people have heard of, but it's the fundamental technology that underpins artificial intelligence. And it's a way to essentially create a model from a set of data using machine learning algorithms. And those models can be used for a lot of very wide variety of things. They can be used to help drive self-driving cars. They can be used in facial recognition. They can be used to predict demand or predict whether machines need to get maintained or they can be used to decide whether people should get loans or insurance. So they're really a very wide application machine learning and AI is considered to be one of the next waves within technology. What I found when we were starting to use this technology was very immature. There were a lot of challenges actually getting the technology to work. And I felt the root cause of a lot of those challenges was the black box nature of the technology and why we call it the black box is because, previously, if you were creating a model—and if anybody's taken econometrics course or anything like that or understands what regression models are, those models of things where you create models out of data, but the human is actually kind of in charge and architecting that model and deciding what features are going to be in the model and exactly how the features of that model will be constructed with machine learning. The value of it is that it's algorithms that are doing what people used to do before. And that allows you to use much, much larger sets of data. But the problem is the resulting model is not something that a human really built by hand. The algorithms, in essence, build the model and the models that they build a very complex. And because of that complexity and the more automated nature of building the model and the very large amounts of data that are used when a model makes a particular transaction or outputs, you don't know why it's done that and you don't know how the inputs to the model relate to the outputs of the model. And this creates a whole bunch of problems. It makes it harder to build models, but it makes it harder for people to trust models. It creates risks like models could potentially be biased. They could learn information from the past because we know that we're living in a world where we're bias exists and these algorithms are trained on past data. And how do you make sure that these models are not going to be biased? How do you make sure that we can use this technology ethically and responsibly? And then how do you make sure that this model and machine learning technology continues to perform over time because the world changes coronavirus and COVID is a perfect example of that. If you, for example, had built a model that was trying to predict car accidents, well, that the model that if you use the data before covid, when people drove a lot, there were lots of traffic jams and people who were commuting—that model would have learned certain things, certain relationships between inputs and outputs. But now that people aren't commuting as much, it's a very different world. And that model that you've built before likely is not working as well. So how do you detect that and change that and solve that kind of problem? So that's the company that we built today. It's based on research from my two co-founders who worked at Carnegie Mellon university, who did six years of research to figure out the mathematics for how to explain these models. So it's built on a technology called explainability, but it does a lot more than the more than that. So that's the latest project. And I think it's both really interesting from a technical and technology perspective. But what I really like about it. And what I really wanted to do in terms of starting this company was doing something that I also felt would be beneficial for the world. There's a lot of fear about AI and I feel that that fear is because it is partially caused by some of these challenges and that we can help people adopt AI in a way that's responsible and ethical and reduces those fears. Very cool. 

LJR: So when you look back to someone who felt that medicine was their calling and then you said, no, I, I see kind of a working world in business, policy in my future, what would you say to the younger Will about where you are now and what would he say about it to you?

WU: Well, I might say broaden your horizons a little bit so you can get to this point a little bit faster. I felt that I had a fairly closeted but, you know, it was everybody grows up in their own environment with their family and the friends and influences. And looking back at it, I wish I had been an engineer and computer scientist in college. I wish I had learned some of the technologies earlier that I'd figure that this out a little bit earlier. But, you know, that's hard without having knowledge of the broader world. So I encourage trying to acquire that knowledge of what everything that is out there. But to do internships or to talk to people in a variety of different fields, and then the one thing that is consistent, though, is that I always wanted to do something that I felt would be impactful and positive for the world, that that was the reason for medicine. Quite an obvious way to help people when there's some form of medical distress. And that, I always felt that that would be a worthy and fulfilling thing to do. Here I feel similarly, but it's a different way. It's more about helping bring new technologies to market that can improve standard of living. That can also be an enabler to solve lots of these problems. So there's a huge opportunity to use AI in health care, for example. And actually one of the barriers to using AI more in health care is, is the exact problems we're talking about the black box nature. And you have to build really safe models and models that work consistently and that can be explained and understood by medical practitioners and also the patients. So have as broad horizons as you possibly can. Talk to as many people as you can about finding what your passion in life is would be the main thing I would suggest. And then I would also just suggest, don't be afraid to take the harder courses in college, have that broaden your horizons from a cost perspective as well. And I look back at some of the things that I did from a college perspective, in terms of coursework and the like, and say, I'm really glad that I did that and that if I had the opportunity to do more, I'd do more of that. And even if it was hard and maybe I wouldn't get as good grades, it would be better to get the learning and get the skills than necessarily to worry too much about grades. 

LJR: Yeah, well, it sounds like you ultimately got like you needed to put you on the track to be where you are. So I think it has been great. Thank you so much for sharing this road with us. It'll be really delightful to see how this new venture takes off for you and where it might lead. So thanks again, Will.  

WU: Thanks very much, Leslie

LJR: That was Will Uppington, CEO and co-founder of Truera, which helps enterprises analyze machine learning, improve model quality and build trust. Find out more at truera.com. Though his first founder role, this certainly isn't his first rodeo in the technology start-up world. He has a started and grown a number of successful companies and products and always tries to do something that is both of personal interest to him and of use to the greater world. He lives in the San Francisco Bay Area with his wife Lauren Currie Uppington, and their family.  // Each Monday we post another full-length interview episode with a classmate of Will's, Lauren's and mine, as we walk the road this 25th year after college graduation. Join us on the journey by subscribing wherever you access your favorite podcasts. Or check us out at RoadsTakenShow dot com. Thanks so much for listening and if you really like what you're hearing, drop us a note at RoadsTakenShow@gmail.com and please consider leaving a review on your podcast platform so that other people may find us more easily! Thanks for joining me, Leslie Jennings Rowley, on another ROADS TAKEN.