Before coming to Harvard, I spent most of my life preparing to become a ballerina – I had studied briefly at the Bolshoi Ballet Academy in Moscow and was going to stay there to complete my training. But my plans suddenly changed when I sustained long-term injuries because of a particularly taxing three months of competitions. While recovering I learned a lot about myself and injury prevention (and ended up doing some more research on this) It’s been a while since I’ve done a pirouette on stage, but nowadays you can find me either running, playing tennis, or occasionally at the gym. Cambridge has a lot of beautifully well-kept tennis courts (even some grass ones!) so I’ve been enjoying playing some rallies with Bryan.
At Harvard, outside of my physics studies I am part of the TDS department at WHRB, which is the student radio station. During the semester I give a show every week, sometimes together with a friend (nationally broadcast at 95.3FM or streamed from whrb.org). I am also part of SCAS, or the Small Claims Advisory Services, which is a student run volunteer group that helps people who can’t afford a lawyer navigate the small claims court system. It’s a good way to break out of the college bubble every week and help people who really need it. It’s given me some invaluable lessons about how the real world works and a humbling appreciation for what I have.
How I Wandered into Physics and Then Machine Learning
I am inspired by beauty and my appreciation for it is manifested in everything I do, from the delicacy of form in ballet to the elegance of proofs and theorems. I love the surprising power of the universal laws that govern everything around us – the foundation that magically holds true in every possible situation. It can describe the real world, from something as simple as an apple falling from a tree to something as counterintuitive as a quantum computer or space-time near a black hole.
I am drawn to physics because it presents a simple and elegant way of understanding how the world works, and this is what drew me to machine learning as well: machine learning is another way of doing what physics does. In a way, the two complement each other perfectly – physics gives you an intuition for how your system works, which informs the machine learning model you want to build. At the same time, machine learning is extremely powerful in terms of what you can do in the realm of physics.
My Time at Dreams AI
I found DreamsAI when looking for startups doing machine learning, but I joined Dreams AI because I was really excited by the people and the nature of the work that they do. I think what initially convinced me of this was the interview process. The take-home problem was extremely open-ended, which is when I first realized that you have a lot of ownership and control over the work that you do here. I had a lot of fun approaching the assignment from different angles and ended up mapping it to an analogous quantum computing problem. In my subsequent call with Adam, he took my ideas even further, showing me how to extend my solution in a way that I hadn’t thought of before.
Conversations like this have been a recurring theme throughout my time here: everyone is so excited and open to talk about everything, and I walk away having learned something new every time. I’ve spent the first couple of weeks here looking at what others have worked on and the sheer breadth and depth of what’s been done is really impressive -- from Luka’s variational autoencoder code that predicts what language the word you are typing in belongs to and visualizes the result in a 3d latent space, to Milan’s stochastic path optimization that can find an optimal path along any arbitrary cost function, to Nat’s Bayesian network structural learning algorithm. It’s been so much fun playing around with people’s code, seeing how well it’s been done, and getting to call or meet with them to talk about everything and ask questions. It’s exactly the kind of environment you want to be in when you get excited by and want to stick your nose into everything.
All the Fun I’ve Had and More!
Everyone has been so friendly and welcoming and kind to me so far! I’ve had a lot of fun getting to know past and future interns and DreamsAI people. I’ve spent the most time with Z and Bryan, and it’s interesting to get to see their perspective of the company from the operations and design side of things, which is different from what a software engineering intern would normally see. There are so many new developments every day and watching everything unfold is really exciting.
We regularly have board game nights, go to the gym together, grab coffee, and just hang out – there’s always something to do. It’s my first time in the UK so I’ve been shown around some of the colleges and beautiful parts of the city. Cambridge is gorgeous at this time of year, and I even got to experience the second half of cherry blossom season. I’ve been enjoying just walking and biking around and sampling all the many coffee shops it has to offer now that quarantine restrictions are lifting. I saw Newton’s apple tree, which was pretty cool, and I’m still surprised by the fact that they have cows in several parks here.
Although I’ve only been here for just over one month, it’s already been such a busy and exciting time. I’ve learned so much and gotten to talk to so many new people. I’m so excited for what’s in store and to keep working at such a great place!
