Transcript
GET TO KNOW US | JANE STREET
Hi! I’m JP and I’ve been an equities trader at Jane Street since 2017.
What’s your day-to-day like?
It’s a little hard for me to describe my average day because things can change so much. Some days I have my headphones plugged in and I’m really focused on making kind of small modeling improvements to existing models. Some days we’re having sort of “blue sky” conversations on the desk, trying to figure out what approaches we want to take next. And some days when the markets are really busy, I’m more engaged in what’s actually happening in trading on the day, trying to figure out if our models are reacting appropriately to data that might be out of distribution.
What’s it like working in Hong Kong?
Asia is Jane Street’s fastest growing region, and you really see that every day on the desk in the pace of innovation and change as we deal with new problems. And that’s part of what makes it so exciting to be at Jane Street in Hong Kong. Outside of work, lots of people know Hong Kong is a great city for traveling in Asia, but lots of people don’t know that it’s also a really outdoorsy city. I’m a big trail runner and there are great hiking trails really close to the office and also beaches less than an hour away.
How has Jane Street’s ML work evolved over the last few years?
While data and modeling have always been central to a lot of the work we do here, the scale of our research infrastructure and compute has grown a lot, and also the breadth of different modeling techniques we use every day is totally different from what it was when I started.
Continuing to scale our machine learning infrastructure is one of our biggest priorities at Jane Street. We’re rapidly increasing the amount of CPU and GPU compute that’s available to researchers, as well as building out large internal libraries for sharing modeling techniques across our regions and desks.
What’s it like using ML in finance?
So there are a few things that make working on machine learning problems in finance different. One is that finance problems tend to be very low signal-to-noise, compared to a lot of the problems that people deal with in other fields. So it’s really important to think about that in terms of your architectures. Another thing is that the market actually evolves against you, and you have to be really careful to think about how things change once you implement your strategy in the market.
What kind of people do you look for to join the ML team?
We’re not necessarily looking for one profile or background for machine learning at Jane Street. We’re always interested in finding smart, curious people who are interested in tackling these problems with us, both on the modeling side and on the infrastructure side. There are lots of different ways to contribute.
What’s your favorite part about working at Jane Street?
My favorite part about working at Jane Street is the collaborative environment. Everyone’s always sharing ideas on the desk, helping each other solve problems, and building toward what we’re going to work on next.
