Start ordering with Uber Eats

Order now

Scalable Systems & Scalable Careers: A Chat with Uber’s Sumbry

February 28, 2018 / Global

What do Site Reliability Engineering (SRE) and mentorship have in common? According to Uber SRE manager Sumbry, both areas focus on growth.

At Uber, Sumbry balances his time between ensuring that our systems grow reliably at scale and volunteering with UberHUE, the company’s black employee resource group. Like engineering extensible systems, he says, mentoring group members also leads to growth and development, but in the context of their careers.

We sat down with Sumbry to discuss how he got into SRE, his passion for mentorship, and his work with UberHUE:  

When did you first get interested in engineering and technology?

I actually got interested in computers from a pretty young age. One day my mother came home with an Apple IIe computer. I started playing around with it and eventually taught myself BASIC. That kickstarted my love for tech. I eventually went to the University of Southern California for computer science and then joined my first startup afterwards—an Internet Service Provider. Since then, I’ve worked for several other tech companies, from a social media platform to a telecommunications startup, and now Uber.

How did you get into Site Reliability Engineering?

Partially by accident. I started off my career as a software engineer, so early in my career I wrote a lot of code and built a lot of software. I quickly discovered that, while I loved coding, I was more interested in building infrastructure and scaling services. Over time, I gradually shifted towards SRE, although it wasn’t called that back then. I think Uber is the first company where I had the SRE title, but I’ve done this work for quite a bit of my career. Whether we called it operations or platform engineering or a subset of infrastructure, it was basically the same thing.

What do you enjoy most about being a site reliability engineer?

I wanted to work on this more challenging side (at least in my head) where I’m not just coding, but I also have to understand the infrastructure, I have to understand the platform. I have to learn how to scale an application from a few requests to tens of thousands or hundreds of thousands. of requests.  There are many interesting problems that need to be solved beyond just coding an interesting feature and the industry is only now just starting to acknowledge them and staff teams of people dedicated to solving these problems.  We refer to them as the non-feature or non-functional properties of a system, and working in that area is a big reason why I pivoted to SRE.

Why did you decide to join Uber Engineering?

I was referred here by a friend who I’ve known for almost my entire career  who I had worked with previously.  I was initially reluctant to join Uber honestly, but over time, I got the opportunity to meet more people at Uber and start hearing about some of the interesting problems and challenges they were working on solving regarding scalability and reliability. After all, if Uber doesn’t work, people get stranded, and that was really, really interesting to me.

I had a bunch of informal meetings, and I was like, whoa, Uber doesn’t seem like the company portrayed in the media at all. Once I really got to meet some people and hear about their experiences, my perception started to change. Then I interviewed more formally, but at that point I was already all in.  

What is most interesting about your work at Uber?

At Uber, we are doing things that no other companies are really doing yet. I think maybe every, like, seven-ish years or so, you get a tech company that comes along and does something completely differently than the companies before it. When I joined Uber, I felt like the industry was really starting to make the shift into logistics. It was really fascinating to me because Uber is not just real-time logistics, it’s also a marketplace that pairs with other interesting areas, like geospatial exploration, data visualization, mapping and routing, and streaming data. You really get to work with a mixture of technologies, techniques,  and industries in a completely new and different way. I literally learn something new at Uber every single day.

In your opinion, has technology become more inclusive since you first joined the industry?

I’ve worked in this industry my whole career, and I definitely see more of a focus on diversity and inclusion recently, but I still think we have a lot more work to do.

You’re a member of UberHUE, Uber’s black employee resource group. How can tech companies create more inclusive environments for engineers?

I don’t know if I have an answer to that… there are just a bunch of things we’re trying. I think maybe that scattergun approach is part of the answer. It’s just acknowledging that when this emphasis on promoting diversity became a focus, it was just “diversity.” At some point the “inclusion” was added, and I think it was because people realized that it wasn’t enough just to hire diverse engineers, but you also have to make them feel accepted.

I heard a good analogy about this once: diversity is like inviting a bunch of people to your house to dinner, and inclusion is actually inviting them all to the table. I don’t think we’ve figured the inclusion part out yet, but we’re at least trying. For a while, most companies didn’t even try.

Over time, we’ll figure out what works well and what doesn’t. However, there’s a couple of easy things companies can do to promote inclusion, and this is true for engineers of all backgrounds. Give them mentors. Give them coaching. Find a way to organize them into a cohort, a community, so people don’t feel isolated and alone.

What sort of initiatives does UberHUE lead to promote inclusion at Uber and in technology more broadly?  

One of the big things UberHUE does is outreach, which is something I’ve been involved with for awhile. Our outreach focuses on building awareness of the opportunities available at Uber. We’ve reached out to a lot of historically black colleges and specifically recruited there. We also partnered with the Hidden Genius Project, an organization that trains and mentors young black engineers for successful careers in technology. Once, when I was visiting one of these colleges, a student actually said to me, “Wow, I didn’t know any black engineers worked at Uber! Our outreach helps dispel this myth.

In terms of inclusion, UberHUE sets up a lot of events for the whole company. For instance, we’ve created a lot of activities and educational opportunities around Black History Month to raise awareness. We organized a group to go see Black Panther. Anything from a formal outing to a social lunch contributes to growing this community and promoting inclusion.  

What role has mentorship played in your own career?

Early in my career, I did not have any mentors. Now, I have a several great mentors here at Uber, and I mentor numerous people at the company as well. I also belong to some mentoring groups outside of Uber, one of which is /dev/color.

I think the reason I’m so passionate about mentorship is because I didn’t have it starting out in my career. A lot of the time, I was stumbling around trying to figure out what to do. If you aren’t around tech your entire life, you might not realize that being a software engineer is a viable opportunity, and even when you figure it out, it’s not always clear what you’re supposed do and learn. Having a mentor can really help point you in the right direction.

Now as a manager, my mentorship revolves around the things I need to do to improve my management and leadership skills. Mentorship is truly more of an art than a science, and there’s no one way to be a mentor. It’s a powerful tool for getting more out of your career and accomplishing your goals.

What’s some of the best advice you give your own mentees?

I try not to actually give specific advice because it may be too prescriptive, and a wise person once said there is a fine line between hubris confidence.  I don’t want to seem like I have all the answers (I don’t) and I actually learn a lot from hearing my mentees’ own experiences.  A lot of stuff happened to me during my career, so I just usually tell stories about myself and what happened to me in certain situations and hope the shared experience and knowledge is useful. I love reading biographies for this exact reason,  because I feel like I can learn from other people, and I think it is much the same with mentoring—just talking through problems and having those great conversations goes a long way.

Outside of your engineering work, what drives you?

I love to solve problems. But you can probably tell that from the way I answered your questions about how I got into SRE. I’m a little bit of an addict in that the more interesting the problem, the more attracted I am to it. And it’s not just technical problems, but organizational and scaling ones, too. These types of challenges are what gets me up in the morning.

What technologies are you most excited about using in the future to further scale our technologies and build more extensible systems?

Uber has what we’re currently referring to as “the dimensionality problem.” At a typical company, you might need your system to monitor one page, which is likely a simple check, e.g., if zero or more than 100 people are visiting my site, then maybe I should alert someone because I’m getting more people than I expected or not getting enough.

At Uber, our problem actually fans out across multiple dimensions. We have all these different cities that we operate in, and then on top of that, we have all these different products, like uberX and Uber Eats and Uber Black. Then on top of that, we have all these geofenced areas, like airports, and then smaller hexagons we use for dynamic pricing and things like that. Monitoring each one of those things means that there’s not one thing we care about, there’s perhaps 5,000 super-critical things, and then hundreds of thousands of things that are slightly less significant, but worth keeping an eye on. In this context, using traditional monitoring techniques doesn’t work. Adopting technologies like machine learning to accomplish our jobs is no longer a nice-to-have, it’s a necessity!

Outside of your engineering work, what hobbies and interests do you have?

I have a lot of hobbies and interests because I feel like, if you’re 100 percent into computing, you’ll go crazy. So I am a DJ and music producer, I’ve been doing that for awhile, and I’ve had the opportunity to travel a lot, playing gigs and producing some tracks. I’m a big movie and TV buff as well. I love sci-fi, but I’ll watch anything, I think because I’m from Los Angeles where movie culture is very pervasive. You’ll often find me in the hallways of Uber chatting about the latest movies in theaters or great shows on Netflix.


Interested in working with Sumbry and his team to architect more reliable and resilient systems? Apply for a role on Uber’s SRE team!