Kevin Novak joined Uber in 2011 as the 21st employee, its seventh engineer. By 2014, he was head of data science at Uber. He speaks proudly about that time but, like all good things it ended. Novak left Uber at the end 2017 having achieved what he desired at the company.He began to pick up the pace of his angel investment, which he had already started focusing on on weekends and evenings. He eventually built a portfolio that included more than 50 startups, including the fintech Pipe, Standard Cognition, and an autonomous checkout company Standard Cognition.Novak also started advising startups and venture companies such as Playground Global and Costanoa Ventures. He was also a member of the Data Collective. After falling in love, Novak decided to open Rackhouse Venture Capital in Menlo Park, Ca. Rackhouse has just closed its first fund, $15 million. It was anchored by Curtis Chambers, Uber's first head engineer; Steve Gilula (a former chairman at Searchlight Pictures and Cendana Capital), is the fund's chief executive officer. Many of the VCs Novak is familiar with are also investors in this fund.Novak was available to talk about the new vehicle when we met up late last week. We also discussed his tenure at Uber. He was a key player in creating surge pricing, though he prefers dynamic pricing. You can listen to the full discussion below, or see excerpts, which were edited for clarity and length.TC: You had planned to become a nuclear scientist. How did you get to Uber?KN: I studied physics, mathematics, and computer science as an undergrad. I wanted to be a teacher when I entered grad school. However, I enjoyed programming and applying physics concepts to the programming space. The nuke department had the most supercomputer time so it was a great place to do a lot of my research. The research that ultimately led to the Higgs boson, which was the funding for my studies to become a nuclear scientist, was very indirect. It was a great discovery for humanity, but it was terrible for my research budget. . .A friend of mine was interested in what I was doing. He knew my skills and suggested that you check out Uber Cabs. It's like a limo service with an app. It was a very interesting data problem. I also had to solve a math problem. I applied, but I wore a suit and tie to my interview.TC: You're from Michigan. I grew up in the Midwest and understand why people might wear suits to interviews.KN: When I got off the elevator, the friend who encouraged me to apply said, "What are you wearing?" But, I was asked to join as a computational algorithm engineer. This title predated data science and I spent the next few years in the engineering and product industry, building data features and. . Things like our ETA engine. This basically predicts how long it will take for an Uber to reach you. My first project was to work on tunnels and tolls. It was difficult to figure out which tunnel Uber used and how to construct time and distance. So, I drove the Big Dig from Boston to Somerville, and back to Logan, with a lot of phones, to collect GPS data.While I learned a lot about Uber cities, my main claim to fame was dynamic pricing. . . It was a cornerstone of the strategy to ensure Ubers were always available.TC: What do you think about that? When you tell people you invented surge pricing, how does that sound?KN: This is a quick test to determine people's enthusiasm for finance and behavioral econ. Wall Street people think, "Oh my God, that's so cool!" and then many people say, "Thank you, yes, thank you so very much," which is a wonderful thing. They will buy the next round. . . [Laughs.]Data was also the incubator space for many of the early special projects, such as Uber pool, and many of the ideas around, "Okay, how do you build a dispatching system that allows different people to pooled ride request requests?" How can you efficiently batch them in time and space so that we get the right match rate [so that this] project is financially viable? We spent a lot time thinking about the theory behind hub-and-spoke Uber Eats delivery model and how to apply our lessons learned about ride-share to food. This gave me the first-person perspective on many of these products. It was three people writing notes or riffing on laptops over lunch. These small businesses eventually grew into large, national companies.TC: You worked on Uber Freight the last nine months, so you were there when Anthony Levandowski's business was exploding.KN: It was a very interesting time for me. After more than six years, I was already developing an attitude that Ive done all I wanted. At the time, there were 20 people in the company. . I felt like I was at a natural stop point, and I sort of missed the small team dynamic. Then Ubers 2017 came along and Anthony was joined by Susan Fowler and Travis had a terrible accident in his private life. His head clearly wasn't in the game. I didn't want to be the guy that bailed in the worst quarter in company history so I spent the next year trying to keep the band together, motivating and empathic, and being good in all senses of the word.TC: You left the company at the end that year, and it seems like youve been busy ever since. This includes launching the new fund with outsiders' backing. Rackhouse is a better name than Rackhouse. When you invested your own money, Jigsaw Venture Capital was the name you used.KN: Yes. KN: Yes. One year before I started investing in angels, I had established an LLC. I was marking my portfolio for market and sending quarterly updates to my accountant, my wife, and myself. This was one of the exercises that was carried over from my training managers. I believe that you can grow most efficiently and effectively if you have a few skills. So, I tried to figure out how to manage my own back office. Even if it was moving money from my checking account into my investing account and writing my portfolio update.I was excited about the possibility to launch my first externally facing fund using other people's money. However, there is a Jigsaw fund in the UK. As I began to talk to LPs, and said, Look, I want this data fund, and I want it be early stage. Then I realized that I would be competing with Jigsaw in mindshare, so I decided to create my own brand.TC: Have you rolled any angel-backed deals into your new fund? Rackhouse now has thirteen portfolio companies.KN: I have a few people that I've committed to moving forward and storing for the fund. We are just going through the technicalities of that right now.TC: Machine learning and AI are the main focus.KN: Yes, that's right. I believe there are incredible opportunities that lie outside of traditional industry focus areas that, to the extent you can find rigorous AI applications, will also be less competitive. Deals that don't fall within the strike zone of as many [venture] companies is what I want to play. This opportunity is available to all sectors, regardless of geographic biases towards domain experts.TC: Does that explain why you want to keep out of the strike area of most venture companies?KN: I want to ensure that I create a fund that allows me to participate in the early stages of companies.Zack Bogue and Matt Ocko [of Data Collective] are close friends. They were mentors and small LPs in my fund, and they talked to me about their journey. They now have billions of dollars in assets under management. The people I [like] to back are two people who are working as freelancers and are getting ready to make the leap. [Firms the size Data Collective] have priced themselves out of formation and pre-seed stages, which I enjoy. This is where I have a lot useful experience. It is also the stage where you don't need to have five quarters of financial knowledge to be convinced if your domain expertise is strong.
