Jessica McKellar
  • Years in Tech


  • Current Role

    Director of Engineering, Dropbox

  • Place of Origin

    Fremont, CA

  • Interview Date

    March 5, 2016

I’m an engineer and startup founder who spends a lot of her time at work and in open source communities improving diversity and inclusive culture. Fun facts: I’m a technical consultant for the HBO show Silicon Valley. For several years I was a Director for the Python Software Foundation, the 501c3 non-profit that stewards the Python programming language and community. For my diversity outreach work in the Python community, I received the O’Reilly Open Source Award.

Let’s start from the beginning. Tell me about your early years and where you come from.

1987, I was conceived, no—where I came from—as it relates to being a quote-unquote woman in technology, the origin story is probably about like this: I was a science nerd in high school. I was really into chemistry. I really liked my AP chemistry class. I was president of the science olympiad team, in Nashville, Tennessee. It was not a school that was particularly known for its academics, to be totally honest. I found my spot there.

I was born in in Fremont, California. There were a few reasons my family moved to Nashville. One of them was the chaos of being a student in the Bay Area at that time. My parents are aging hippies and they just weren’t that into exposing their kids to that level of insanity. I sort of did my own thing for high school and I seemed to turn out okay, so I appreciate that. I was a science olympiad nerd. I applied to a couple of different schools. I got into M.I.T. to study chemistry.

That’s actually my first degree, I have a Bachelor’s degree in chemistry. While I was pursuing a chemistry degree, I had a bunch of friends who were pursuing computer science degrees. I would look at them out of the corner of my eye and see what they were learning. It seemed pretty different. Compared to what I was learning it seemed like they were learning a toolkit for solving arbitrary problems in the world in a way that was really cross-disciplinary. I felt I was learning the history of chemistry in a way that could be really fun for some people, but what they were doing seemed very broadly appealing and relevant. I ended up taking a couple of CS classes. I actually sort of illicitly pursued a computer science internship without telling my chemistry advisor. He was quite unhappy with me when he found out about that. I ended up liking it a lot, and I ended up getting a bachelor’s degree in computer science as well. I got a master’s degree in computer science, and the rest is history.

That’s how I got started. I did take a couple programing classes in high school, although I wouldn’t claim that I understood the power of it then. I really started pursuing the discipline in college. I think that’s actually important to note because a lot of my friends at school had been programming since they were babies. This meant that I had a lot of catching up to do. This instilled in me a very serious and personal empathy for beginners. I also had to validate for myself that if you work hard, and if you work harder than anybody else you can catch up.

So I did that. As I got my master’s degree, I had some friends who were starting a startup called KSplice. KSplice’s story is the gift that keeps on giving, like the Silicon Valley, like the HBO show Silicon Valley has derived a number of entertaining stories from KSplice the start up. We were all computer nerds together at MIT’s computing club. They started KSplice around a technology for rebootless kernel updates on Linux. It’s like a very exciting, serious, distance program, a challenge.

“The good people and the smart people are going to treat you equally and like a human being and not like a woman in technology.”

They were a little bit older than me so when I graduated, I joined them as an early kernel engineering hire at the startup. We built it into a real business that ended up making money, being profitable. We did a bunch of goofy stuff during that time. This is like the authentic startup experience where we were working at the house where all the founders lived. It was a disaster. Dubiously zoned. We’d all have to hide when the landlord came by. There were mice everywhere. Dudes were showering with their girlfriends during the workday. All kinds of goofy stuff that has made it into various Silicon Valley episodes. It was a really good experience in learning what it means to build something from scratch.

Totally self-funded. Didn’t take any outside investment. We worked all the time, didn’t pay ourselves for a long time. That company was acquired by Oracle, and we hung out at Oracle for a little while. We left Oracle, went back to one of our apartments and continued to startup number two with the same team. Startup number two was called Zulip, it was building a real-time collaboration service for businesses. Sort of like Slack before Slack was Slack. Did the Zulip thing for a while. Zulip was acquired by Dropbox about two years ago. Now I’m at Dropbox where I’m an engineering director and a chief of staff for a VP of Engineering.

How was your experience as a female entrepreneur?

The good people and the smart people are going to treat you equally and like a human being and not like a woman in technology. The people that I was very intentionally surrounding myself with, like my peers at MIT and then my co-founders, are awesome and we work well together and we trust each other. I would take a bullet for those guys.

There was never an issue. I wasn’t the face of fundraising because my role as a founder was to run engineering organization. Our CEO and our COO were all fundraising while we were busy building the product.

The first time in the context of my professional career that I had a rude awakening about this: We had this really, really successful blog. We got really good at writing blog posts that we’d get a bunch of attention from the tech community. We knew how to get to the top of hacker news. We once made the mistake of having a blog post that included a picture of us. That is when I had the great pleasure of receiving a lot of commentary about how I looked in a blog post that was, otherwise, about engineering.

That was maybe the rudest awakening. There were also a lot of people misgendering me when talking about the author in the blog post, even though my name in Jessica. That would be a small cut in a death by a thousand papercuts situation. I don’t really try to put my face out on the internet, especially on things that are closely associated with my social media, because it’s not that fun.

I always felt really good about running a startup and the people I was surrounded by. Now at Dropbox, I am the most senior woman at the company currently. I believe I have a tremendous amount of freedom and flexibility and respect. I feel good about that, but again, I think it’s who you surround yourself with that matters. There are plenty of assholes that you could maybe end up being surrounded by, but the good people and the smart people, they know what’s up. They know what actually matters is having great people, totally respective of their gender or other qualities.

How did you end up as a technical consultant for Silicon Valley (the TV show)?

The show is this wonderful uncanny view of the valley and what’s uncomfortable about this industry. I only worked it tech for a few years, but I made it through one episode and it was like “This is too real” and I couldn’t get through. Like I said, Ksplice is really like the gift that keeps on giving. So what happened was that the Silicon Valley writers they were doing a tour of a bunch of the startups in the valley and they came to Dropbox.

Our PR team was like, “Who are some entertaining people in engineering?” and my name came up. They shoved 25 people into a conference room to ask questions. They wanted to just hear ambient stories that could be interesting and they wanted to dig in a little bit on storage, which is part of a plotline for season one.

I seized upon the opportunity to regale them with many amusing anecdotes from Ksplice. It was all kinds of goofy stuff. The service we were building is a technology for rebootless kernel updates on linux. You care about rebootless kernel on linux if you have a lot of computers and if you have a lot of computers frequently you are a hosting company or you are in the porn industry.

I just have so many stories around debugging technical issues on the computers of various companies in the porn industry, and they are just very funny. I thought that if it could tell a bunch of these goofy stories to the Silicon Valley writers and they thought I was sufficiently entertaining that they wanted to keep talking to me so we kept meeting.

They would ask me funny questions and I would tell them stories. I ended up becoming a senior technical consultant, which is important because James Cowling, who is another person from Dropbox, who is a technical consultant for the show, he’s just a technical consultant. So my name is in every credits and his is only in the episodes he directly contributed to.

That was fun and we- James and I together- helped design their little data center from season two. They invited us to the taping for scene one of episode one of season two, which is filmed at AT&T Park. They had all the celebs out, the Winklevoss were there. Drew and Raj made the cameos in their goofy stuff. It was fun to watch the behind the scenes take on how this stuff gets made. It was fun.

I love it. Let’s go into some of the other things. Aside from your job at Dropbox, you do a lot of work on improving diversity and inclusive culture at work and in the Python community. Tell me more.

I started using Python in school. It is one of the primary languages that many IT classes are taught in. I used Python in all my internships. I’ve actually used Python in every job I’ve ever had.

My first ever contribution to an open source project was while I was an intern at VM Ware. I was using this library called Twisted, which is a Python library for event-driven networking in Python. I had noticed that some of the documentation was confusing, and I thought that this might be an opportunity to contribute back to an open source project. I heard that was a thing that you could do. It seemed scary, but maybe I could figure it out.

I decided I was going to do it, but I was super nervous about it. I quadruple-checked all of the new contributor guidelines for this incredibly simple documentation badge. I agonized over the ticket creation, the title of the ticket, and the body of the ticket, and hovered my mouse over the submit button for multiple minutes, and it was like, “Someone is going to yell at me. This is going to be terrible.” I finally hit the submit button, and then, they’re like, the nicest people.

Glyph, who is one of the creators at Twisted, who is now a good friend of mine, personally helped me through getting this first patch applied. He is super patient and helpful and nice. He is not at all intimidating, even though he’s the creator of the project, and had every reason to not help me. Despite being incredibly nervous, I had this super, super positive first experience with open source contribution, which is not what most experience.

Anecdotally, that is not a typical. I think it ended up being good for Twisted, because I ended up contributing a lot more to the project. I ended up becoming a core maintainer for the project and writing a book on Twisted with O’Reilly. I really invested in the community over the years after that initial contribution. It also made me very aware of the fact that this is not an experience that everyone necessarily has. It made me very committed to wanting other people to have that experience as well. That has become my personal direction open source.

That was happening and then in sort of a parallel work stream there was this program that was happening on the West Coast while I was still in Boston as a student. It is called RailsBridge, which is like one of the earlier efforts in this wave of diversity outreach, to get more women into the Ruby/Ruby and Rails community.

I got to talking with Ashish Leroya, a friend who is also an open source nerd too, about doing something like this in Boston. We didn’t know Ruby. We knew Python. We thought, “What if we just do this in the Python community?” So we hooked up with the Boston Python user group and convinced and Ned Batchelder.

We convinced this guy Ned, one of the long time organizers of Boston Python, to let us run an intro to Python workshop, specifically for women in Boston, under the Boston Python User Group. It was sort of a bold thing to do because I didn’t really know who we were. Getting the right messaging out can be a little tricky, but he was game for it. We ended up selling out, basically, immediately. It was free, but it ended up filling up immediately.

We ended up running a bunch of these. I became an organizer for the Boston Python User Group. Then the Boston Python User Group, in large part due to these types of initiatives and other very intentional initiatives around providing a more inclusive user group for people of various backgrounds.

“As it turns out, if you make an effort to be welcoming to one set of people, you will probably actually become more welcoming to everybody, which is the secret to all diversity, everybody just actually helps everyone.”

Boston is a great testing ground for this, because it has a very high density tech population. It is adjacent to a bunch of other stuff, like bio-tech, a lot of entrepreneurs, music is very big—so there are a bunch of adjacent fields that leverage programming that you can tap into if you do a good job of being welcoming.

We ran a bunch of these diversity outreach workshops, and then a bunch of crazy statistics occurred. We set out a goal – it’s as true in my day job as in these open source communities – like you set a goal and then you measure it to know if you’re like actually doing what you intended to do. We had this goal around increasing the representation of women at user group events. We started at nearly 0% at user group events for women, wanting to boost that to 15%. We instantly achieved that by running these intro workshops and some follow-up events. It really became a pipeline of events.

Then we were able to sustain that participation rate for several years. I’d have to check in with Ned about what the latest stats are, but while I was there in Boston, this represented a huge leap in the actual composition of the user group but also the way that it felt. It became a real poster-child/example set of processes that were adopted by user groups around the United States and globally. We gave a talk about this at PyCon, and it’s been replicated like all over the world, which is pretty amazing.

So I was an organizer for the Boston part of the user group, which is part of this initiative, became the largest user group in the world. As it turns out, if you make an effort to be welcoming to one set of people, you will probably actually become more welcoming to everybody, which is the secret to all diversity, everybody just actually helps everyone.

I then ran for and then became a director for the Python Software Foundation, which the non-profit and stewarding organization behind the Python programming language and community. I served as a director for a couple of years and I was also a co-chair for our Outreach and Education Committee, which provided a lot of funding to educational initiatives. They use Python.

For the last three years I’ve been the diversity chair for Python. There are many Python conferences around the world. Python is the original big international conference that’s held in North America somewhere once a year. I’ve been the diversity chair for the past three years, so in the three years I’ve always had that role formally.

There’s a pretty sick set of statistics about this. If you look in the old Twitter feed, like three years prior to me having this role, the percentage of speakers at Python who were women was one percent, and then it was 5%, and then it was 10%. When formalized, and I was investing time in and getting other people to invest time in, it was 15 %, and then it was 33%. By the time this project launches, that number will increase to 41%, which is insanely high for a very prestigious open source conference. This is a conference with a highly competitive and actually mostly blind selection process, so this is a pure top of the funnel investment with real payoff.

You’ve engineered a system that was successful in increasing diversity in this specific culture. What are the things that have worked and what are the roadblocks that you see that make it tougher to replicate across all tech?

I should caveat the whole prior discussion with the fact that gender is not the only demographic that matters. Gender happens to be relatively easy to measure, in a way that’s not super creepy, and track over time. Women are half the population. It’s a very obvious needle to move. Let’s pretend that we wish there were more women in engineering in our company. It can be women, or it can be any other demographic that you’re trying to optimize for. Here’s the process:

Step zero: it has to actually be a place where people want to work. For example, it has to be a place that equitably retains and promotes women. There’s no point worrying about and investing a bunch of time in the hiring process if once people join, there they’re going to be unhappy and they’re going to quit. That’s step zero. You have to measure this stuff or you won’t know if you’re doing a good job. Do that.

Step one: You have to have an equitable evaluation pipeline.You’re the only one who’s measuring it, so you need to measure segments broken down by the demographics that you care about. If your evaluation process is fair, and has equitable outcomes, this will stabilize across the demographics that you care about. There’s a standard pipeline analysis for this, which is like your pre-onset to onset, and your onset to offer, offer to offer accept. You have to ensure that whole process is equitable.

“If all it is a math problem, what it really is, then is just a prioritization question because anybody can do the math. It’s a question for you as the head of a company, what is the appropriate way to prioritize this problem? If your company is running out of money tomorrow and has a bunch of other problems, maybe this isn’t the thing that you need to be paying attention too. But I suspect if you want to be a company that’s around for the long haul and you want to attract the best talents and retain it, you probably have to care about this whether you like it or not because people like me are going to opt out of working at your company if you don’t make it a priority.”

The beauty of this is that when you achieve steps zero and one, it becomes a numbers game. You have to pass through raids on an evaluation pipeline that is equitable and dumps candidates into an environment that is equitable, all you have to do then is have some outcomes that you want, back up the math on that and that’s a set of top of funnel targets that your job. Your job is to incentivize a diverse top of the funnel that will cause the math to happen from the pipeline that has those outcomes. That’s all it is. It is just a math problem.

If all it is a math problem, what it really is, then is just a prioritization question because anybody can do the math. It’s a question for you as the head of a company, what is the appropriate way to prioritize this problem? If your company is running out of money tomorrow and has a bunch of other problems, maybe this isn’t the thing that you need to be paying attention too. But I suspect if you want to be a company that’s around for the long haul and you want to attract the best talents and retain it, you probably have to care about this whether you like it or not because people like me are going to opt out of working at your company if you don’t make it a priority.

Then, it’s a prioritization question, so you make sure that it’s staffed appropriately with recruiting and with event coordinators and it’s incentivized appropriately with your hiring managers and with recruiting. Then you just do it. Any company that says that they care about this, in particular, if you’re going to write a blog post about it or something where you’re going to say that you care about it and then you report back next year that nothing has changed, you don’t have to make up a crazy story about it. You just didn’t prioritize it sufficiently. Just own that. It’s your decision to make and you can evaluate the pros and cons, but that’s all it is. It was a prioritization decision. Once you have an inclusive culture, it will make for an equitable pipeline and a properly incentivized top of the funnel and that’s all there is to it. So, simple, right? How do people screw this up?

Well, it’s funny hearing just the very first step and making it—even just for tension it’s just starting to be talked about. It’s crazy.

If that sounded depressing, the good news is that I think that it has become something that people have to care about and it will just naturally become a thing that accompanies our better add over time. Enough people are going to be doing this. Kids these days who are going to  graduated from college soon. They’re going to enter into a work force, or they’re not even going to realize that it used to be difficult to get engineering teams to talk about why this was a thing they should care about. It’s a different scene these days. That’s extremely encouraging. We should be vigilant about this stuff. It’s easy to regress, but I’m pretty optimistic.

Do you feel the inclination to apply your knowledge, and what you’ve learned, and your experience, into different specific niches of tech after this? After Python?

That question is, what do I actually care about in life, which is a totally different question, although we could talk about that, if you want.

We can talk about that.

I am not strongly motivated by being in an engineering leadership position at a tech company. I appear to be pretty good at it, which is why I keep doing it, and then keep getting into increasingly large levels of responsibility for it. I really like people. I really like surrounding myself with the types of people that you find  in these organizations. There are other things that I actually care about in life and I would expect that in the arch of my adult life and I will move into a pretty different role over time.

Things that I care about include—they’re going to all sound related. Things that I care about include the democratic machine, like democracy as an institution mostly in the United States. I care about education. I care about journalism. And then I care about the power of media more broadly to educate and influence people. These things are all related. You need an aware and educated population that is able to work together in a democratic society to move each other forward. That’s how it’s all related.

All of these things, one property that’s really nice about being in a software company or being in software, is the scale. You can write software that you continually distribute to everybody, to hundreds of millions of people if you want. That is incredibly powerful and once you have a taste of doing that, it can be difficult to do work that scales less. For example, it is difficult to visualize myself being a teacher in a classroom. Maybe if I’m a teacher in a classroom that is on the  internet and reaches lots and lots of people, maybe that’s appealing. Or if I’m helping to drive policy decisions that impact many, many people, that’s appealing. It’s finding the impact at scale in these fields that I’m pretty passionate about in a way that parallels the opportunities scale that you can have in software. Those are things I actually care about.

Why have you been working in tech companies and for a while if it’s not really what you care about?

I’ve learned that it’s been intentional. I’ve learned a bunch of highly transferable skills building companies and managing increasingly complex organizations.I wouldn’t go back and undo it. I will eventually, over the course of my life, move into something that’s more differently related to one of these four topics that I mentioned. A change from holding a leadership position in a pure tech company.

On that note, I’m curious about your high-level thoughts on the state of tech in 2016, and what excites you, specifically the potential of tech’s applications to those things that you care about.

It’s a good question. I mean some of these things are a little obvious, and if I were better predicting the non-obvious things, like maybe I should be a VC and just cash a bunch of checks, ha!

I’m a pretty avid Twitter user.  When I watch grassroots political movements on Twitter that’s exciting to me. The internet is amazing. It connects people independent of distance. That’s why the internet is the greatest invention of all time, and I’m a little bit obsessed with it. There is also the infrastructure of the internet, we could go off on a whole weird tangent about that. That’s exciting.

To follow up on that, that was a lot of the stuff that I think is the biggest set of issues isn’t going to get solved. Sometimes the solution isn’t technological. There will be technological platforms that will enable the people who are going to cause the change to do it, but at the end of the day the thing that’s going to make a difference. Problems like fulfilling the promise of a equitable  public education in the U.S. is not going to be solved by technology. I don’t think it is—I could be wrong. Sorry to any ed-tech people who end up reading this.

I don’t think like a piece of software is going to solve that problem. It’s going to be the policies and the institutional infrastructure around people that we change that solves it. Which is part of why I probably won’t be building the software for very long. I’m going to be more directly in the sector with the people who are making the change. Probably.

Do you have feelings about technology influencing the rate at which policy can be changed?

You would hope that it makes it faster. If you could vote from your phone, or a computer, that’d be pretty great. The place that I just was before I was here, I was just in Buenos Aires, and they have compulsory voting. Because you have a bunch of people who vote because they have to, and they don’t know what’s going on. There are pros and cons to compulsory voting. It’s too bad that the engagement is so low in the U.S., but if we could make it easier to vote, that’d be pretty sweet. Although in this election, I think that we’re going to have better turnout than ever before. It’s a crazy time.

What is frustrating to you about tech in its current state, and what would you like to see change?

I think it’s bad that in the US there is a monopoly on this sort of startup ecosystem and culture. Really the only game in town is Silicon Valley, and I’m sort of a walking example of this. I was in Boston, started two tech companies in Boston and even we succumbed to the pull of Silicon Valley.

It’s because money is better out here. The tech companies who are going to acquire you are going to pay bigger premiums out west. The talent’s all out here. The network is out here. And that’s bad. What do we learn in history about monopolies? Like, monopolies are—in the long run they’re bad for innovation.

I would love to see truly competitive alternative to Silicon Valley in the United States. I would love to see competition that it’s a healthier culture, and maybe promotes people being willing to invest in ideas deeply, for a long time. Really bringing expertise in a domain to solving real problems in the world, as opposed to just building apps. That is my fantasy is that 20 years from now.

From all of your experiences, having not worked only in Silicon Valley proper, being a female developer, doing tons of diversity work, I would just be curious to hear your general thoughts right now on the state of diversity in tech. Aside from the system that can fix it, or even the high-level roadblocks you’ve seen.

If you look at the research, the data tells us that there’s a leaky pipeline and that diversity for women in industry as compared to graduation rates is still trailing behind and it’s a much worse story for other demographics. If we wanted to pick on ethnic demographics, the story for African-Americans and Hispanics is much worse. That sucks! I don’t want to be in a different industry though.  

Why do we even care that the tech industry be reflective of the population that it serves? It has such a profound ability to change people’s lives if we do it the right way. I personally wouldn’t want to leave because of the shitty pipeline issues that I’d rather fight, but not everybody has the luxury of being able to do that. I mean, the trends are weird, right? The trends for women in CS were going down for a really long time. My understanding is that they’re back on the upswing, but the peak was decades ago, and we have a lot of ground to recover. I feel like there’s a lot of work to do, but that’s work worth doing.

I keep going back to the story you shared of having such a positive first experience in shipping that patch to the open source project. I wonder how massively important those first experiences are, and whether they’re positive or negative.


—and whether or not that turns a person into a incredibly high contributor vs someone who eventually leaves the industry.

Anyone in data tells us that it’s usually important. It doesn’t even have to be anecdotal, I think we just know from the data.

This is related to some of the things we’ve talked about, but it’s worth saying. I get asked quite a bit why I spend a bunch of time on outreach in Tech, and the real actual reason for this is different from what people would expect. I believe we will build better products if the people who are developing these products are a  reflection of the population that we are serving.

It’s the right thing to do. I want any kids that I have to feel like they can do anything in the world. Those are all the reasons why, and I believe that it is important that people have access to programing.

“Learning how to program allows you to develop the confidence that you can learn how to operate within a system and to deconstruct it and to tear it down to make it better.”

Not everybody has to be a programmer but let them always have the opportunity to try it out. It is the first time that you can become fluent in a system, so that you understand that you can change it and that lesson that everything in the world is a system that can be understood and deconstructed and changed in that way. I think many people probably never have had the luxury to navel gaze about this stuff. Learning how to program allows you to develop the confidence that you can learn how to operate within a system and to deconstruct it and to tear it down to make it better.

Knowing how to do that and have confidence in that as a programmer is already very powerful because when you know how to program you can create all of these amazing things in the world. The more important lesson there is that you have gained—you have had this experience of observing a system and believing that you can change it. And that’s the thing that I actually want everybody on the planet to experience because there are many systems, and some of these systems are software systems, but probably the most important ones are people systems, and it’s not actually any different.

Everything is a system that you can understand, and you can deconstruct, and you can break down, and you can change.”

Everything is a system that you can understand, and you can deconstruct, and you can break down, and you can change. That is the thing that I actually want everybody to experience. That is why I spend all of my time—or a bunch of time—teaching beginners of diverse backgrounds how to program, because that is maybe the most profound thing that I’ve ever experienced, and maybe one of the more profound things that many, many people in the world can experience if they have a chance to do so.

What advice would you give to folks who hope to get into programming or are just getting started?

I’m a lot more confident now than when I was when I was 18.I don’t know if there’s a way to short circuit that. I don’t know if there’s a world where anybody can tell you like believe in yourself, have the confidence. I don’t know if there’s any way to learn that through experience but it makes such a big difference. Given my experiences in college, feeling like I had to work harder than all of my peers who’d been doing this for a lot longer than me, and later to get to the point where I could prove myself to them and validate that I was as good as everyone else like a big validation.

By having the confidence to dive into that and to totally believe people actually thought I was good wasn’t simple. It was like an imposter syndrome. The side issue I want you to learn about is imposter syndrome. It actually makes material impact in your life. This is one of those beautiful things that you realize that if you put into it you could reflect on it could actually change your behavior. I think that’s how I would summarize it. If I had known about it when I was 16 or 18, that would have been helpful.