Stefan Li
Can you introduce yourself with your name, year, major, and hometown?
I'm Stefan Li. I'm a third year here in Berkeley studying CS. And I'm also trying to get a Film major here, but it's all spontaneous. I am from Beijing, China.
How did you choose your majors?
As a freshman, I was unsure what to study. I'm still trying to find out which niche of the tech field that I want to be in, going into the second semester of my junior year. But in terms of majors, I came in as a math major, then took on CS quickly. But I later realized that math upper divs are really abstract. And I was kind of unsure about what I wanted to do with this math major if I graduate with it. The skills I was gaining, if I don't go for a PhD, won't be much too useful in the future, and my love for the beauty of mathematics sort of faded away as time went on. So I am more focused in CS now. As for film, I've always been a cinephile and watch movies every week; it's fun.
What are your hobbies and interests?
Besides film, I'm a big basketball fan. So I came to Berkeley partly because of the Warriors. I was choosing between a couple schools on the East and you know, Berkeley, I think basketball was my final punch of the decision. Yeah, so I play and I watch. And I play video games. Honestly, not much besides film and basketball, just studies. I did play the guitar, I was in a band in high school, but I didn't have time once I came to Berkeley.
What professional experience do you have? What are your professional interests?
I've had three internships in the tech field. The first one—it's funny—I worked as a machine learning intern as a freshman, for a medical startup called Airdoc, on a team of all PhDs. I learned image recognition from scratch. It was fun! At the time, I was also debating whether to pursue a PhD. That experience pretty much put the PhD idea to bed, because I felt like work would widen my scope and present me with more opportunities, whereas a doctorate dives deeply into a subject. Since I wasn't sure where my passion's at, I decided that's probably not what I wanted. The second internship was done in Tencent, mother company of WeChat. They are building a video sharing platform to compete with TikTok in China. And I worked as a data analyst. The analyst role turns out to be a potential career route that I think I can thrive in. Thirdly I worked as a Software Engineer, building machine learning infrastructure. It was the most challenging out of the three internships. It was a demanding team, but I gained a lot from my mentor, who would literally read all my code and give suggestions.
Are you involved in any other campus organizations?
I used to be very involved in TOPPA. "TOPPA" has similar pronunciation as the word "breakthrough" in Japanese. So as a club, we aspire to motivate East Asians in tech to break through their career "glass ceiling" by getting a head start. We did projects together, held workshops, and brought in mentors from Google. I was one of the founding members of the club and was VP of Marketing until junior year.
How did you find out about SAAS and why did you join?
Because of my friend Linda, who was the Web Dev director for last semester and the semester before that. She just kept ranting and ranting on about how good this club is. So I mean, I gotta check it out. And I was lucky enough to get into one of the DC teams a year and a half ago.
What committees have you been on, and what did you do and learn in each one?
This is my third semester in SAAS. So the previous two, I've been in a DC team working on, firstly, an image recognition assignment. For this first one, I think that was my first exposure to real-world data science projects. My biggest takeaways are perhaps how the client focuses on different things than, say, our professors do on the data science project. They are looking for not only accuracy, but also speed and how they can deploy the projects. I think for me, I did a lot of extra work that the client didn't need; I think maybe communicating more ahead of time would be something I would do if I got involved in another project. I would probably do more of that, which is exactly what I was trying to push in the second project, which was working on a time series prediction on ProducePay, a company that sells produce. We were trying to predict the shipping point prices of American produce, which I think is more applicable. This project was actually at the core of their company, so we got to talk with our client a lot. They had this very formal presentation and dozens of engineers who showed up. They were impressed with us, but they also made a lot of suggestions. I think that was the highlight of the project for us, presenting to the core of the company.
This semester, I'm in EVP. I'm curating a faculty panel. It's not going very well, so I don't know, we'll see. The timeline kind of got messed up and different professors, they have different timetables and they don't necessarily match up. Last semester was a coffee chat, organized by Christina, and it was just one professor at a time, and honestly only a few students showed up for each coffee chat. But we are looking to either make the event a bit more formal, so that people show up, or not do it at all.
Why did you decide to stay in SAAS?
SAAS is just a great hub of community and opportunity. Honestly. We're very diverse. That's what I like, a lot with people that are not only diverse but people who also share the same interest in data science or CS or stats. It's just great to be in; there's opportunities everywhere and friendly people like yourself. I've made numerous friends over Zoom, though unfortunately not in person yet for this semester and the last. I love it. What's not to love about SAAS?
How has SAAS helped you with your career and professional development?
I went to a couple of workshops when I first joined SAAS; I think the first one was a course selection workshop. But I think the most memorable one was, I can't remember who organized it, but it was a networking workshop. I've always been an introverted person, and networking has been pretty tough. I think the workshop really helped me to approach strangers and form a conversation. Apart from that, just the #opportunities channel. There's a lot of jobs that I actually want to learn about and companies that I would like to learn more about. I think recruiting is not like a linear assignment, it's not like you do it for three months and you're done. You have to learn about different companies as you go on, different people as you go on, and form a relationship and try to maintain contact with them. So I think the #opportunities channel really gave me another channel to explore the companies and explore different opportunities.
What is your proudest accomplishment in SAAS?
I think the model we worked on in the ProducePay team is my biggest accomplishment so far in SAAS. So, we were re trying to predict prices for the American produce. I was deploying a time series framework called Facebook Prophet. We were able to predict three weeks into the future, with a margin of error of less than $1, for all four species of produce included in the presentation. If we were given more time to automate hyper-parameter tuning, perhaps we would be able to train all the species of produce (several thousands in total). I know it's probably incredibly difficult to implement, but the thought of having such an ubiquitous, robust model is an exciting idea.
What are some ways in which you've been able to make friends with people in SAAS?
Retreats are great. My closest SAAS friends were acquainted during the retreat I went to. Hopefully in-person retreat can resume soon. We rented a country house, partied a bit, and played Avalon. Also there was this one game called "human knot", where we held each other's hands to form a rope and tried to untangle — it's more fun than it sounds; definitely try that. Another way of building relationships is working together on projects and spending time together as a team.
Of the three SAAS values of community, exploration, and mentorship, can you choose one and explain what it means to you?
I'll talk about exploration. The environment we are in greatly influences our decisions, and Berkeley's tech vibe definitely plays a huge part in my choice of career entry point, which lies in Data Science. SAAS really widens my scope of the DS and CS realms — I see potential roles I can grow into in the community built by SAAS. A lot of the alumni and advisors like Brian and Ronnie were really helpful and gave me insights so that I'm more comfortable with the situation I'm in.
What is your favorite class?
This question has so many dimensions. Well, I'll give you a couple, and perhaps come to a conclusion. Data 100 is pretty fun. A good class needs a good professor; a good class with good content will not be a good class without a good professor. So Professor Joseph Gonzalez, his instruction just really fit my way of thinking, and that really helped me throughout the semester. We were learning practical skills, obviously, which is rare for a lot of CS courses being so theoretical at Berkeley. I also liked CS 170, which is algorithms, which has a reputation of being not very likable at all, it's a very difficult course. But I have a passion in mathematics and the algorithms that we learn or like the essence of mathematical ideas that we were able to come up with in a couple hundred years, that was just a fascinating course to take in. That was also useful, because it helped with interviews, coding interviews to be specific. So I invested a lot of my time in it and looked over different algorithms online on my own. Mostly, I wouldn't do it for any course, because we tend to fill up our schedules just by the courses. So these are the two technical courses that I really like. I am taking a neuroscience class this semester, called Psych, or Molecular Biology, I think they share the codename C61. It's really interesting that the professor's too cool to be true, like he goes to Tibet to teach monks and priests neuroscience, and he also goes to Nepal to go into the forest and take mushrooms for his studies. David Presti. So he talks about marijuana, stuff like that in class, which is just fascinating. One of the most interesting classes I've taken. So I didn't come to a conclusion, but any of those three classes.
Where is your favorite place to study on campus?
So two scenarios: if it's not dead week, I do like my apartment because it's really cozy. It is facing west, on the seventh floor. I can see the sea and I have like eight windows in my bedroom. It's the corner room, so sunshine is the best. But if it is dead week, I will go to Main Stacks because I can just concentrate there.
In your opinion, what are the prettiest and the ugliest buildings on campus?
Prettiest is Evans. Ugliest is its twin brother Evans. Yeah, that's my formal answer; won't change. Nobody will sway my opinion.
What is your favorite restaurant in Berkeley?
I think I'll go with Chengdu Style. Because it took me a couple months to really get used to the cafeteria. I was in Unit 2 and Chengdu Style was always on my way back to my dorm on southside. It opens until really late and they do takeout. And I know their owner personally, so that shows how many times I went there.
What is your favorite Berkeley memory?
There are so many. It's an almost impossible question to answer! But I'll go with the time my family came over during sophomore summer and I was showing them around campus. When we came to Upper Sproul, I told them how Mario Savio emphatically gave the "put your bodies upon the gears" speech from the stairs of Sproul Hall in 1964, at the peak of the Berkeley civil rights movement. In 2020, I think people are starting to realize that America is going through a cultural identity crisis. And at that moment, I truly realized I'm part of the Berkeley legacy, which stands for the most admirable qualities of America. I'm just grateful for being here and sharing the campus with so many brilliant minds and kind-hearted people. At that moment, I was the most proud about myself and Berkeley.