What are you working on right now?

My background is in machine learning, and I mostly focus on applying it to accelerate scientific discovery. Right now, I’m a member of the technical staff at EvolutionaryScale, a company spun out of Meta AI that built one of the first protein language models. I’m leading a research project around foundation models for biology — using AI to make discoveries in science.

How did you get into machine learning?

I started coding in school through robotics, then spent a lot of time teaching myself — reading papers online, following tutorials, and building projects.

Being in Toronto helped a lot. I remember going to a University of Toronto conference and meeting Geoffrey Hinton. He told me to go home and learn backpropagation. I didn’t know what that was at the time, but it sparked my curiosity. That moment nudged me deeper into ML, and I started building based on what I was reading.

What was your first project?

In high school I built my first product at 15 or 16. It focused on supply chain transparency for counterfeit medication. Developing countries like India and Pakistan face huge challenges with counterfeit drugs, and I saw blockchain as a tool to track authenticity.

I hacked together an MVP, reached out to companies like IBM, and ended up with engineers mentoring me. Some of what I built even made it into their code base. That was an early lesson in how far a scrappy build + cold outreach can take you.

What excites you about the work you’re doing today?

There’s a big shift happening in AI for biology. Most people hear about AI for coding or dev tools, but there’s incredible progress happening in life sciences.

At Ark Institute, I’ve been contributing to a virtual cell model. The idea is: can we simulate a cell well enough that instead of running expensive wet-lab experiments, you could ask the model? For example: “If I perturb this protein, what happens?” That’s a game-changer for how science gets done.

Have you thought about starting your own company?

Yes. Before college, I actually took a gap year in SF through a fellowship and worked on a company called Bolt X. We were using ML to speed up battery testing. We did pilots with companies, raised a pre-seed round, and learned a ton.

I realized I needed to become much more technical. Having ideas is one thing, but being able to build the MVP yourself gives you independence and momentum. That experience pushed me to go deeper into technical research so I’d never be limited by not being able to build.

What advice would you give to someone trying to get more technical — especially self-taught?

Projects > classes. Classes give you theory, but real learning comes from doing.

A few things helped me:

• Professors and grad students: Don’t be afraid to reach out and offer to help. I learned a lot just by working on their projects and writing terrible code until I got better.

• Community: Find people to learn with. I ran informal biology reading groups in high school where we’d read papers and discuss.

• Posting online: I shared projects on GitHub and Twitter, and that drew other young builders to me.

You don’t need permission — just start building, failing, and sharing.

Where do you find inspiration?

Mostly from friends. Having peers working on cool problems is the best fuel. Just talking with them about what they’re exploring sparks ideas for me.

I also read a lot of papers and do deep dives when something catches my eye. Twitter helps sometimes too, though it can be an echo chamber.

What technology excites you most over the next five years?

I’m biased, but I think AI in science is going to be transformative. Yes, there’s hype, but if you focus on foundational improvements — making the core models and methods better — the downstream applications will be enormous. Especially in biology, we’re only at the beginning.

What advice do you wish someone gave you early on?

Balance exploration with depth.

When you’re young, it’s good to try many things. I’ve done that, and it’s helped me figure out which problems I really enjoy. But at some point, if you find something that excites you, double down. Go deep. Don’t stay in the middle forever.

The same applies to starting a company: explore, but once you know, commit.