Thuan Pham promised himself that it was time to take a vacation. He spent more than eight years helping build VMWare’s engineering team, seeing the company go from 300 to 15,000 employees. Pham needed a break before figuring out his next career step.
But then Travis Kalanick came calling. It was late 2012, and the Uber CEO needed an experienced engineering guru who could help manage the underlying technology for one of the fastest-growing companies in the world.
“I planned to take a year off,” Pham said. “But I didn’t make it.”
GeekWire recently sat down with Pham for an interview at Uber’s Seattle engineering office, where the ride-hailing giant employs 100 people and plans to double its workforce in 2016. Pham has an inspiring personal story, arriving in Indonesia more than three decades ago on a refugee boat from Vietnam before starting life in the U.S.
After graduating from MIT, Pham worked at HP Labs, Silicon Graphics, DoubleClick, and VMWare, where he held leadership positions from 2004 to 2012. He was then recruited by Uber board member Bill Gurley, who helped set up a meeting with Kalanick.
Since Pham became CTO, Uber’s valuation has skyrocketed (now at more than $60 billion) and the company now operates in more than 400 cities across 68 countries. Just like any growing company, Uber has had its share of engineering-related hiccups, but Pham has helped build sophisticated infrastructure for the company to run smoothly amid the huge growth.
“It’s the hardest job I’ve ever had — but that’s why it’s the most fun job I’ve ever had,” Pham said. “Every day I get to do new things, and every day I feel like I’m getting better in some way. And that’s because of the gnarly problem we have to solve.”
Read on for more from our conversation with Pham, which details his interview process with Kalanick, the inner workings of Uber’s engineering team, and his advice for budding entrepreneurs. The interview has been edited for clarity and brevity.
GeekWire: Thanks for taking the time to chat, Thuan. How did you arrive at Uber?
Uber CTO Thuan Pham: “I was at VMWare for eight years. When I joined, the company was pretty small, maybe 300 people. When I left, there were 15,000 people. As the company got bigger and bigger, we went through a hyper-growth period and I got a taste of what that was like. We saw lots of good things and lots of bad things and I try to apply both the mistakes and the successes to my job here at Uber.
After a while, when VMWare plateaued in terms of saturating the market, it was becoming boring. I am very adverse to getting bored because you don’t learn anymore. You have to consistently put yourself in an uncomfortable situation so you can keep learning. My last two years, I was staying there for the team I led. We had nearly 1,000 people on the team and I was captain of that ship and didn’t want to bail. Then there was a CEO transition and the direction was changing anyways, so it was a perfect time to leave. I planned to take a year off after that, but I didn’t make it.”
GeekWire: OK, how did Uber get on your radar?
Pham: “When we announced that I was leaving, a bunch of people asked me to look at things. Uber board member Bill Gurley asked me to look at Uber, to spend one hour with [Uber CEO] Travis Kalanick. If I didn’t like it, I could always say no. But of course, who wouldn’t like Travis?
I met with him, and that one hour turned into two. He canceled his subsequent meetings, which seemed like a good sign. He invited me back for more talks and ended up interviewing me for 30 hours straight, one-on-one, over two weeks. It was on Skype, too — he was traveling around the world at the time. This is how Travis does it.
He sent me a photograph of all the topics he wanted to talk about. That was a result of our very first meeting in his office. We’d pick each topic and drill all the way down. One topic could be how to hire and fire people. Another was about project management and engineering management. It was any topic related to engineering, and we’d just pick one and debate it. I had my view and he had his, and because he’s an engineer by training as well, we just jammed like that.
Interestingly, we didn’t always agree, but we always managed to get to a mutually-acceptable decision at the end. It dawned on me afterward that he wasn’t looking for someone that shared his view — he was looking for diversity of thought, someone who could challenge him and who he could challenge. So that’s how we got to the best answer. It was cool.
Throughout those 30 hours, it wasn’t about getting hired. I actually forgot it was an interview. It was just like a discussion between two colleagues. I got to know how he thinks, what he cares about, what he’s passionate about, what he hates, etc. Finally during a Sunday morning session, he just stopped — we hadn’t finished all the topics — and said, ‘I’ve had enough. Let’s talk offers.'”
GeekWire: But you were supposed to take a year off.
Pham: “I wasn’t ready. But of course, Uber and Travis move at Uber speed. He said, ‘we have to do this.’ I told him I needed 30 minutes. I went online, did some research, put stuff in my head, and called him back. He then broke off for an hour and came back and we had a deal. Everything moves super fast at Uber. Once we decide to do something, we just lean in. Then the magic started.”
GeekWire: You could have gone anywhere. Why Uber?
Pham: “Throughout my journey, I discovered three things that cause me to work anywhere. One, is there a big mission? I don’t have many years left in my career and want to make every year count. I saw Uber had potential to have a huge impact. Second, do I like the team? You spend more time with each other at work than our own families. And I really liked the team at Uber — smart, motivated, passionate people. Third, I have to like my boss. With Travis, I could sense what kind of a world-class person he was.
I thought that if I didn’t give myself this opportunity to work with him — whether it turned out good or bad — I might regret it. So, I said that my one-year vacation thing could wait. And it’s still waiting.”
GeekWire: That was 2013. Let’s fast forward three years later to today. How much have things changed from what you expected?
Pham: “At the time, neither Travis or myself realized how big this could be. Because it moves so fast, the only thing we could really do is look out six months to a year and figure out what we needed to do to get this thing growing.
When we launched uberX, it exploded. Then we got into more cities and we had to get more efficient. On the engineering side, we had to build a system that was scalable and do it at a fast enough pace so that it was reliable while the business was growing on top of us. It was breathless. Everything we touched kind of turned into gold, but it required tremendous hard work and personal sacrifices. Up to this point, anything I had envisioned, this has far exceeded it. No company has ever grown this fast.”
GeekWire: Let’s talk Seattle. You opened an engineering center here last year in order to support that insane growth.
Pham: “We didn’t want to start until we found an awesome, strong local leader. We finally found Tim Prouty, and just look what happened one year later. We have 100 world-class people. We thought we’d maybe have 50, so this is beyond our expectations. That’s just the way the pace of the company works.
Seattle, of course, is important. There are so many things that are great in this market. Silicon Valley is the mecca of everything with Google and Facebook, but Seattle is the next thing with the Amazon and Microsoft. But even beyond the big companies, there are lots of real innovative startups and the awesome University of Washington system. There are lots of great systems people, OS people, and great computer scientists. The kind of talent base here is very, very amenable for standing up a world-class engineering organization.”
GeekWire: The Uber app seems so simple to use. But the engineering underneath the skin of the app is complex. How important is engineering to Uber?
Pham: “Clearly, without the technology and engineering this thing wouldn’t work at all. But it’s half of the equation. The other half is operations, the men and women managing the physical marketplace and dealing with the political environments and everything else. The two major turbines of the company are engineering and operations, and we power this airplane. We work pretty tightly together in terms of unifying behind the mission, but at the same time we are separate enough that everybody can move super fast and not completely gang up on each other all the time. We just build features, roll it out, and the operations side basically takes it and makes magic with it. It’s a symbiotic relationship and it’s pretty amazing.”
GeekWire: The pace seems to move really fast at this company.
Pham: “We move faster than other companies. It’s brutal hard work with a gnarly problem we have to solve, but somehow we get it done and we look back and think, ‘how the hell did we do that?’ It’s those moments when we push past our red line and we look back and realize our range is actually bigger and we’re more confident in our abilities than three or six months ago. We keep pushing our own personal boundaries and accomplish things we didn’t think were possible.
If people were to join our company for the wrong reasons, like because this is an ‘IPO ride’ or because it’s a hot company right now, they at best will be miserable because they did not sign up with the right mindset. It moves super fast. When we launch something, we are building the next iteration right away. You have maybe a week to catch your breath, then it’s back on the treadmill again. I now look at my friends at VMWare or Facebook or Google and everybody seems to move in slow motion compared to us.
Uber is actually a similar type of company to Facebook and Google. You hear a lot of people say Uber is ‘just an app,’ but it’s actually 95 percent systems. The app is just the skin. There are servers, dispatch, dynamic pricing, supply positioning, mapping, routing data — all of that infrastructure. Saying Uber is just an app is like saying Facebook is just a Timeline. There’s so many systems underneath that it’s like an iceberg — most of it is underwater that you don’t see. It’s tremendously challenging and also fun to work on.”
GeekWire: You have more than 1,200 engineers now. How does the company get work done?
Pham: “Even though we went from 40 engineers to 1,200 in three years, we don’t have any bureaucracy that comes with a bigger company. We organize ourselves into small, mission-oriented teams — it’s like a startup within a giant startup. We have 50 or 60 of those feature-oriented teams and each has a non-overlapping mission. One will solve a safety issue, one will solve customer support, one will solve the driver and rider app experience, one will solve mapping and routing — you name it. Each of those things has a really clear mission and has an assortment of engineers, data scientists, product managers, designers, etc. They just run with it and have their own roadmap, their own execution plan, and they just ship it. They don’t really tangle themselves with anybody else. Then we have a dozen infrastructure teams — storage, compute, reliability, etc. They enable every team to move super fast by sharing common architecture and services.
The magic is managing that structure. It’s like a matrix structure — where is the accountability? How do you make sure you hold people accountable when the people you manage are actually working on some other project? We’ve managed to crack that code. We just have to do extra work as managers to be aware of people’s contributions to have a good way to assess impact. But those mission-oriented teams is where the product delivery happens.”
Pham: “The thing I learned at Uber that actually made me a far, far better professional was the appetite for risk-taking and not being afraid of failure. Look, when we grow this fast, we’ll never do it perfectly every step of the way. We stumble, we fall; we make really big decisions and really small decisions, and some don’t work out. I’ve made big decisions that completely failed and we had to recover. But with a company growing this fast, not making any decision might be even a bigger failure. So you have to be fearless in charging forward, being on your toes, but also being very grounded and understand what problems need to be solved with what priority. You have to have the courage to solve problems, even if you don’t have all the information or all the answers. You just have to use all your experience and knowledge and battle scars to make the best decision you possibly can. If you are really good, most of those decisions will be right. Some of them invariably will fail and not work out and be very painful. But you have to just charge forward and do it.
The key is, when you fail, do you get up quickly? Do you learn and fail fast? Do you get better right away from that learning? As you charge forward again, you charge forward as a better person with a better team that is more knowledgable and equipped to face the next challenge. We’ve had outages and all kinds of trip-ups, but all those result in us being smarter as a company. The quicker you fail, the more you learn, and the better you become if you get to survive and grow.”
GeekWire: Your company is rolling out new services and products beyond the traditional ride-hailing platform. What’s the future of Uber?
Pham: “We can’t predict it, but we have a vision. Today we are serving single-digit millions of people every day. But there are seven billion people on the planet. Facebook already has more than one billion daily active users, and it isn’t even in China. That tells us they already have a billion people using smartphones with Internet access and everything else. Imagine if we eat into that and get to serve those people and move them and things around cities around the world — imagine how big we could be.
This is just Day 1, Month 1 of this journey. We want to bring this experience to people around the world and improve cities. Can you imagine if we can move 100 or 500 million people around the world every day? How many cars could we take off the road? How much congestion can we reduce or eliminate? The world will be completely transformed through what we do. We firmly believe that.
But, to be honest, we didn’t have that ambition three years ago. We were trying to make 30,000 trips a day work. But when you get into this landscape, we’re starting to see massive changes to people’s lives already, especially on the driver side. People now use us as a platform for on-demand work, for them to be their own boss, an office on wheels. On the rider side, we bring this amazing, fluid, seamless, beautiful experience that is also economical at the same time.
When you can move food, lunch, people — anything — we become that operating system between the digital and physical world. Our mission as a company is to provide transportation that’s as reliable as running water, everywhere for everyone.”