Kevin Miao is graduating this spring with a master's degree from UC Berkeley's Department of Electrical Engineering and Computer Sciences. (Photo/ Josh Edelson)

When Kevin Miao arrived at UC Berkeley from the Netherlands, he wanted to study biology. Taking Data 8, an introductory data science course, helped Miao find a different path. He was introduced to coding and gained confidence in that skill. He took on leadership roles and helped students who might feel othered in technology spaces.

Last spring, Miao graduated with a bachelor’s degree in computer science. Now, he’s graduating again with a master’s from Berkeley’s Department of Electrical Engineering and Computer Sciences and soon will join Apple as a machine learning engineer. 

In this Q&A, Miao describes his journey into computer science, the barriers he faced as a first-generation college student and his passion for building community for people with underrepresented identities in these spaces.

This interview has been edited for length and clarity.


Q: You’re about to graduate with a master’s in computer science, but when you arrived at Berkeley you wanted to major in biology. Can you share how you got here?

A: I'm originally from the Netherlands, and I got into medical school there. But I decided to go to Berkeley. I was way more interested in biology research. So in my first year, I worked in a biology lab. But I noticed that there was a lot of coding to be done.

For that reason, I took Data 8. That was a really interesting journey. Data was so powerful and so useful that I got drawn into the major. The major was friendly, but in the beginning, it was also intimidating, right? Realizing, “Oh, I’m international, first-generation and never heard of coding/ academia before.” It was kind of daunting, but I still felt supported.

Q: So you switched to be a data science major, but you graduated with a degree in computer science. Why did you switch?

A: That was a really big step. The most interesting thing to me was data's machine learning component. I worked at UC San Francisco as a research assistant on biomedical datasets, so I did a lot of data analysis. What I realized was that a lot of these algorithms are not transparent and don’t inform you on how it makes its predictions. 

I started thinking about problems in equality. How do we ensure that these models are fair to different groups of people? We see a lot of health care discrepancies across different races and ethnicities, and people from different socioeconomic classes. Also, I think there's another component: safety. For example, it's fine if your phone doesn't recognize where your eyes are, but if an algorithm incorrectly predicts your healthcare outcome, it's really disastrous. 

That's how I started going down that rabbit hole, like, “What if we start creating and thinking about a new paradigm of machine learning?” That drew me more into the engineering side. But I think at the end of the day, it's still an intersection between computer science and data science, ensuring the safety of all these algorithms.

Q: You mentioned that, coming in, the idea of coding was intimidating. Why was that?

A: First of all, I'd never coded before. And when you first come into Berkeley, people always talk about doing coding camps in high school, maybe even in middle school. It's very scary coming in, sitting next to those people and realizing that you had never done something like that before.

Also, I had less representation. I didn't meet as many people who had that same background as me, which kind of discouraged me from taking Data 8 or other courses until my sophomore year. In the beginning, I was like, “I'm never going to major in computer science.”  But a smaller atmosphere in CS 88 [Computational Structures in Data Science] and the welcoming atmosphere in Data 8 helped me ease into it. Especially in my Data 8 discussion section, I met a few people who were like me and had never coded before. We're still friends today, which is really awesome.

Q: What aspects of your identity are you referring to when you refer to “people like you”? Those without coding experience and those who are first-generation college students?

A: Yeah, exactly. Those are people who are kind of underrepresented in that larger group. I’m also an immigrant. When you think about coding, you think about a specific kind of person, and I never identified as that. I think that’s one of the big inhibitors towards people entering the major. 

Q: How did your identity affect your experience in data and computer science after you’d committed to these fields?

A: I switched my research to focus on making machine learning more inclusive and safe – looking at the downsides of machine learning and how we can make sure that no one group is disproportionately affected by that. 

I also started teaching. I have been a member of the Data 8 teaching staff since 2019 because the classes are inclusive and it's kind of a gateway. This is where students come in – transfer students, freshman-year students – and that's kind of the game, right? If we lose students there, then you lose them forever. So that's, I think, why Data 8 has always held a special place in my heart. I know a lot of people who go on to teach more advanced classes, but I've always wanted to be involved with Data 8. I'm lecturing this summer, too. It's kind of a fun way to end my time here. 

Also, Professor Ani Adhikari asked me when I began my master's to teach a seminar in Discovery Scholars – the crossover program between the Data Science Discovery program and the Data Scholars program. I developed a new curriculum based on what helped me believe I was going to succeed in data science: building community, feeling like a data scientist and getting used to jargon. A lot of people in the Bay Area do use a lot of jargon, like neural networks, machine learning and cloud computing. That was daunting because I was like, “What are all these terms that people throw at me?” So my curriculum was also rooted in talking about it with each other and saying, “It's okay if we don’t know something because we are here to learn from each other.” That helped the students a lot. 

Q: What did you learn from teaching that seminar?

A: I realized the power of data science. In popular media these days, data science is sometimes seen in a negative light, where it's like, “Oh, you know, people monetize on it. Artificial intelligence is taking over the world. People do this and that.” But working with these students, I've learned that there's so much diversity in this field. These datasets can actually solve so many societal problems. 

One student worked on how to reduce incarceration using data science. Another worked on environmentalism, and another researched police brutality. That was really powerful to see the diversity of problems where data science can actually help our society. 

Also, seeing how resilient these students are was powerful. I came from a biology background, so I did have some math background. But some of these students came in from a complete humanities background. Hearing and learning from their stories is so powerful.

"That was really powerful to see the diversity of problems where data science can actually help our society."

Q: You've maintained a really deep involvement with data science, even as you've shifted into the computer science field. Do you find data science informing your computer science work?

A: It goes two ways. My data science work inspired me to make sure my work is understandable. I feel like computer science work, because it's so theoretical, it's very inaccessible to people not in the field. Data science is really accessible to people from other fields, because it's very applicable. So I also try to explain my computer science research in layman's terms when I talk to people, so people understand it. 

Computer science work informs my data science work, too. Essentially, the line between data science and computer science is blurry because they inherently need each other. Computer science provides the technical backbone, the math, theory and infrastructure, while data science delves more into the application, the data, the nature and the diversity of real-world problems.

Q: Any advice you’d offer other students?

A: Don't be scared of something because you don't know it, or people say it's scary. Always try out different things. 

I've taken so many different classes – linguistics classes, economics classes, engineering classes. Because of that process, I learned that I love data science. I love working with data applications and machine learning and thinking about how we can further that for applications that we care about – for social good, but also for improving algorithms that we use in daily life. 

So yeah, take classes that you normally wouldn't take. It's okay to know that you dislike something. But also don't be shy. Just go a little out of your comfort zone.