People of Systems & Architecture: Margo I. Seltzer (Part 2)

With most systems and architecture conferences taking the online route, we figured it’s a great time to get to know a few people in the systems/architecture research community. People of Systems & Architecture is a series of interviews conducted this year, and continues in the same vein as the People of PL, People of POPL, and the People of Language Design and Implementation interviews done by John Wickerson, Jean Yang, Brandon Lucia, and Minjia Zhang.

In this two-part edition, Akshitha Sriraman meets Margo Seltzer, who is the Canada 150 Research Chair in Computer Systems and the Cheriton Family chair in Computer Science at the University of British Columbia. Her research interests are in systems, construed quite broadly: systems for capturing and accessing data provenance, file systems, databases, transaction processing systems, storage and analysis of graph-structured data, new architectures for parallelizing execution, and systems that apply technology to problems in healthcare.

She is the author of several widely-used software packages including database and transaction libraries and the 4.4BSD log-structured file system. Dr. Seltzer was a co-founder and CTO of Sleepycat Software, the makers of Berkeley DB, recipient of the 2020 ACM SIGMOD Systems Award. She serves on Advisory Council for the Canadian COVID alert app and the Computer Science and Telecommunications Board (CSTB) of the (US) National Academies. She is a past President of the USENIX Assocation and served as the USENIX representative to the Computing Research Association Board of Directors and on the Computing Community Consortium. She is a member of the National Academy of Engineering, a Sloan Foundation Fellow in Computer Science, an ACM Fellow, a Bunting Fellow, and was the recipient of the 1996 Radcliffe Junior Faculty Fellowship. She is recognized as an outstanding teacher and mentor, having received the Phi Beta Kappa teaching award in 1996, the Abrahmson Teaching Award in 1999, the Capers and Marion McDonald Award for Excellence in Mentoring and Advising in 2010, and the CRA-E Undergraduate Research Mentoring Award in 2017.

Professor Seltzer received an A.B. degree in Applied Mathematics from Harvard/Radcliffe College and a Ph.D. in Computer Science from the University of California, Berkeley.

[This blog post is a continuation of the interview published last week.]

A: How did you go about establishing your own startup? What advice do you have for academics who want to start companies of their own?

M: My husband and I started the company together when we both had day jobs. We described each week of that first year as the first 40 hours and the second 40 hours. It was kind of crazy. We also had our first kid shortly thereafter. So again, not things I would plan for, but it worked out great. The company was fabulous and I love my son, so it’s all good. 

To have a successful company, you need a good idea, a lot of hard work, timing, and luck.

I think that it’s extremely easy to underestimate the importance of luck. In some ways, timing and luck are tightly coupled. I have little tolerance for people who attribute their success to strategic decisions they made. Yes, they worked hard, had some great ideas, and tried to make good decisions. But, if they had not been lucky, it really wouldn’t have happened. 

I feel fortunate that things worked out well for us. The reason to start a company should not be because you want to get rich. I think that’s a really bad idea, because most companies fail. For us, we had an idea for something we wanted to build that we thought people would find useful. It was the lure of the luxury of trying to build something that we thought would benefit people.

We were building a product for engineers, so we were working with people just like us, and that suited both our personalities really well. I was excited about building something cool that people would use. I couldn’t imagine being yet another app developer. I don’t think that would excite me unless it were an app that I genuinely believed was going to make the world a better place. If I’d been the right kind of person to work on COVID contact tracing, I would have found that appealing.

A startup is not for everybody though. My children grew up with arguments about executive decisions at the dinner table. This probably wouldn’t be in the top ten list of my parenting tips: One of the cruelest things we ever did to our children was telling them that we were selling the company before we were able to tell our employees. We had to sit our children down and explain that the information we had discussed at the dinner table was really secret information that they couldn’t talk about. At the time, we didn’t realize how stressful that would be to an eight year old. I wish we had handled that better, but it worked out okay. The day that the news went public, you could see the weight of the world lift off my son’s shoulders. 

I would encourage you to do it if you’ll be sad that you didn’t try.

Do it because you can’t imagine anything more fun than working on the idea you have. If it meets those criteria, then go for it. We’re in a field where the downsides are: “Oh, so it might not work”. And then, you can just go back to being a professor or get another job. The downsides of our field are just not that bad [pauses]…except, you know, building the next thing that leaks people’s data. Now, *that* would be bad.

A: What are some of the biggest challenges you faced in your career, and how did you go about overcoming them?

M: Remember that luck thing? I was pretty lucky. [winks]

This is going to sound really trite, but I think it’s figuring out when work just isn’t the most important thing and family is. And being comfortable with that.

I remember a conversation with someone who said, “Well, don’t you want to be a Dean?”. I said, “No, I’ve got two kids under the age of 10. Why would I do that?” And he sort of smiled. He said that that was the right answer. Had I chosen a different path when the kids were little and really tried to work my way up the academic ladder, I could have been something else now and I could possibly be, who knows, the Dean, the president, whatever. I’d probably enjoy those jobs, but it would not have been worth the trade off. 

I think the biggest challenge was figuring out how to be both the parent I wanted to be and the professional I wanted to be. Unfortunately, the thing with being a parent is that you get to find out how well you did only after 20 years or so; there are no report cards. There’s not even a paper acceptance. I look at the kids now and I think they’re pretty awesome. So I feel okay. But, you could imagine a different outcome where they might’ve wanted nothing to do with me, and I wouldn’t feel okay about that. 

A: What do you like the most about your job?

M: It’s probably the interpersonal interaction. COVID has been tough because I’m a people person. I could have been an engineer at a company since I really like building systems software. But, the whole reason why I do what I do is because I also like to build systems of people. In many ways, I find both of these system building projects equally enjoyable.

A: You’ve been a part of the USENIX community since 1989 and have held various leadership positions. How has the systems community evolved over this period? What are some stark differences you observe compared to when you attended your first systems conference?

M: Okay, I’m going to sound like an old lady complaining about things now! When I attended a systems conference in the late eighties or early nineties, the systems community there had both academics and practitioners, i.e., people who delivered products. That really appealed to me because I’m a pragmatic person. I had worked in industry. I used to joke that I tend to not try to solve the problems that start with “If pigs could fly…”. I’m more like the person who says, “Show me the flying pigs”. That community really felt right to me and I liked it. So, that was my chosen publication venue for a long time, because it was where you found people who were building real systems that had impact.

You could talk to people about the real problems they were trying to solve.

Ironically, in our effort to make USENIX conferences more well-respected, we ended up making them less relevant to practitioners.

When you see industry involvement in conferences today, it’s almost always researchers from industry as opposed to product people from industry. The biggest transition I’ve seen is that the academic, research, and product communities used to be the same. But now, they’re very different. There is little reason for a typical software developer to go to an academic conference. I think there’s just too big a gap there. I feel like we’ve lost something in that respect. At USENIX, we struggled with this for a long time. But, I do feel that we’ve lost something by that separation.

A: How do you think we can bridge this separation?

M: That’s a hard problem. SIGMOD has a big industry presence too. But, I think the onus is mostly on academics. Most academics feel that whatever industry does is not sufficiently novel. In program committees, when you have industry papers that might take a dozen different known ideas that have been put together in a really interesting way, the papers never got in. Now, sometimes they do, if in fact they can demonstrate that they’ve had ten million users. But, more broadly, I think it gets down to the question of what counts as research.

I feel like we’ve gotten a little bit narrow in what we think counts as research in a way that excludes people who are building interesting products. We spent a lot of time at USENIX trying to fix this and reproduce what we viewed as the success in the FAST community or the other conferences that encourage industrial interaction. But, it is unclear what the right approach is to solve this problem.

A: What is it like being a woman in the academic systems community?

M: In many ways, I feel lucky in that it never occurred to me until I was fairly senior that it was a problem being a woman in CS or a woman in systems. I partially attribute this to my upbringing. We have this joke in my family. My two older brothers were both second in their class in high school being beaten by the “smart girl”.

I grew up *knowing* that girls are smarter than boys and that I had to be the smart girl because our family had lost twice.

It was not a question. It was not a goal. It was just the truth. It had never occurred to me that there was a gender problem, which in retrospect I feel really lucky about.

When I got to college, there weren’t that many women in CS, but I didn’t notice that because I’d always been a tomboy. I hung out with the boys. In fact, when one of my teaching fellows said, “You know, every time I try to talk to you, you’re surrounded by a bunch of guys”, I was like, “Yeah, they’re my classmates. What’s the problem?” So, I proceeded mostly obliviously. In fact, I remember when I got to grad school, there was a special group for women in computer science and engineering, and I was like, “Why?!”  I was happy that they gave us free food once a month, but I literally couldn’t figure out why we had this group.

I was taking a standard grad theory course and walked out of it one day and turned to my office and said, “Wow, there are a lot of women in that class”. My office mate looked at me and asked, “How many? Six?” I said, “But, that’s 20% of the class because there are 30 people in that class”. He was like, “Is that a lot?” I responded that I was taking the operating systems class that just had two women (including me) of 90. That’s when I noticed the discrepancy in the numbers.

As a junior faculty member, I would attend gatherings where senior women faculty would tell the junior women what tenure was about and how things worked. I would immediately share these strategies with my male colleagues. To me, “I was getting all this good info while they weren’t”. Long story short, I was clueless for a really long time in a very privileged way. Then, as I became senior and started to see things happen to my junior colleagues, it didn’t seem quite right. That’s when it really struck me and I was like, “Oh my God! We have a problem!” [slaps forehead].  I then started looking back at things that had happened to me in hindsight and realizing that they wouldn’t have happened to me if I had been a man. 

I vividly remember this incident that happened during my first job outside of college. This was back in the day, before companies fed you twelve meals a day; they would typically just have a big coffee maker. The coffee was usually pretty horrible, so a bunch of us got together and bought a little Mister coffee pot to have our own coffee. A new guy joined the company, and we invited him to join our coffee circle. One day, he sticks his head in my office and says, “Oh, we’re out of cream”. And I reply, “Oh, I do the coffee, Alan does the cream”. He comes back later in the day and says “Oh, we’re out of cream”. I was thinking, “We just had this conversation!” I say, “I do the coffee, Alan does the cream”. He said, “Well, how can I bother him with that?” [?!!!]. I was stunned … “You just bothered me twice!” I didn’t think much more about it. Later that day, I was walking down the hall and he was coming the other way. He said, “Oh, could you get me a cup of coffee?” I looked at him, turned around, walked into my manager’s room, told him what happened, and told him to make it stop. The good news is that they did make it stop (even though this was back in 1984). Somebody actually sat down and talked to him and said, “You’re not allowed to do this”. I took a couple of things away from this experience. One, you have to speak up even if it’s stupid and small and just a cup of coffee.

You’ve got to speak up, because if you don’t speak up about the little things, such as the coffee, it’ll then be about the project and who is running it. It snowballs from there. 

Whether you are the person who’s being mistreated or you see somebody else who is, you have to speak up. If you’re the person being spoken to, then you darn well better do something about it.

It’s as much a part of your job as writing the next line of code.

I feel fortunate that the person who was my manager did the right thing. They sat him down and said, “Dude, you’re being a jerk. Don’t do that”. And this experience pales in comparison to, “Well, if you want the next promotion, you’re sleeping with me” (which also happens).

As I started seeing these things happen to junior people, I was suddenly like “Oh no, no, not on my watch!” One of the things that upset me the most in my last couple of years at Harvard was when I learned that the spouse of one of my students had been sexually harassed while she was a graduate student. From my perspective, this happened under my watch. It wasn’t anybody in my department. I didn’t know the person either, but as a faculty member, that was under my watch. I took that personally. I had a responsibility. I wish everyone took these things personally.

I’ve had black students who face a whole set of other issues. As a community, I feel like we don’t talk about these systemic issues as much as we should. We pretend like their experience is the same as ours, but it’s not. Not one of my white colleagues has ever been stopped by the police because they happened to be in the wrong part of town at the wrong time. My black colleagues have. I am a woman and underrepresented in this field, but there are lots of other ways where I am totally included and feel comfortable.

I’ve started to realize what a privilege that is and how much I take it for granted. When I look at my diverse group of students, I wonder, “Do they actually feel included here? Do they feel like they belong?” That’s important to me: that they feel like they belong in our lab. I think we’ve created an environment where everybody feels like they belong, but I wonder whether they feel like they belong in the department, or the university, or the city? I don’t know. And the fact that I’ve never had that conversation, that’s on me.  We live in interesting times, and I’m trying to take advantage of that and figure out how I can do better. So many of the students I work with are marginalized in one way, shape or form, whether it’s by their gender, sexual orientation, socioeconomic status, or race. I think we need to be more cognizant of that. 

A: I have personally seen how you consciously make an effort to create an inclusive space for everyone in the lab. What advice do you have for someone who is starting a lab of their own and wants to establish a good lab culture?

M: I want to give credit where credit is due. Radhika Nagpal, who was my colleague for the last 15 years at Harvard, single-handedly did more towards helping me understand the importance of establishing good lab culture than anyone or anything else. She is someone who talks the talk and walks the walk. Her lab is both extraordinarily successful and extraordinarily inclusive and supportive. So, in large part, I watched what she did, and I stole her ideas on how to make my lab an inclusive environment.  

There are two other people in my former department, Harry Lewis and Barbara Grosz, who would never turn away a student who was facing a challenge. So, it never occurred to me that you could be any other way. As one of the grown-ups in the room [as an aside, says, “although I hesitate to use that word”], you have to be willing to go to bat for your students when something is wrong. I actually wrote a blog post about this. What is the responsibility you have towards your students? If you’re not going to go to bat for them, why are you in academia?

In my mind, that is the minimum bar: your job is to advocate for your students. 

At UBC, I am responsible for building a big systems group. So, I want to be extremely thoughtful about creating a good culture. I stole a bunch of ideas from Radhika and spent a lot of time thinking about these things on my own. I volunteered to serve on our committee on diversity and equity. I learned a ton from the queer support group on campus. I remember the day when they came into my office and we were talking, and I confessed that I didn’t think I’d have to put my pronouns on my website because it’s obvious what my pronouns are. As I was saying it, I realized that my speech was coming from a position of privilege. I ended up putting my pronouns on my website. Just realizing the fact that “I’m a part of the norm,” makes it harder for someone who doesn’t want to be called “she/her”. I learned a lot from my students. 

I try to listen and stay on top of things. We have had incidents in the past where people have behaved in a way that made other students uncomfortable. When I know that, I take action. It’s not about telling them that they’re bad people. But, just making them realize that we don’t do this here. Those are definitely hard conversations.

I would like to shy away from them just like anybody else, but I realize that I cannot because it would directly be in conflict with my responsibilities as a professor.

If you’re uncomfortable having the whole conversation, I’d simply speak to the student in question and let them know that “We don’t do this here”—this simple phrase can go a long way. 

A: What would you be doing for a living if you weren’t in CS?

M: A soccer player. Assuming I had the talent. I am a huge fan of the US women’s national team and I’m now a huge fan of the Canadian team. I had been rooting for Christine Sinclair to break Abby Wambach’s goal scoring record for a year and a half!

A: Thank you so much, it was wonderful chatting with you!

M: Nice talking to you too!

Bio: Akshitha Sriraman is a PhD candidate in Computer Science and Engineering at the University of Michigan. Her dissertation research is on the topic of enabling hyperscale web services. Specifically, her work bridges computer architecture and software systems and demonstrates the importance of that bridge in improving the performance, cost, and energy efficiency of modern hyperscale data centers. Sriraman has been recognized with a Facebook Fellowship (Distributed Systems), a Rackham Merit Ph.D. Fellowship, and was selected for the Rising Stars in EECS workshop.
She hopes to enter academia after her PhD program, and will be on the academic job market (for tenure-track faculty positions) this upcoming cycle.

Disclaimer: This post was written by Akshitha Sriraman for the SIGOPS blog. Any views or opinions represented in this blog are personal, belong solely to the blog author and the person interviewed; they do not represent those of ACM SIGOPS, or their parent organization, ACM.