Transcript

Get to Know Us | Jane Street

Hi. I’m Alok. I started at Jane Street in 2020 on the research desk in the New York office.

Do you need to have a phd to work in the research group?

I certainly hope you don’t need a PhD to work on research because I don’t personally have one. And in fact, I think the majority of researchers at Jane Street don’t have PhDs. That being said, there is a lot of overlap between people interested in academia and people who come work on quant research at Jane Street. But I think that has more to do with the kind of problems people enjoy solving. I think, in fact, if you ask around, a lot of people would say that instead of doing a PhD, we’d actually love for you to come here and just learn on the job.

What kind of projects do researchers work on?

The biggest category of projects for researchers would be data analysis. This can include the entire end-to-end process of coming up with strategies, starting with building a data set, and then going to look for patterns, training models, and really thinking through the details of how to do a trade that makes money. Other important projects involve building tools. For example, a researcher might identify a general problem that is being solved in multiple places and decide to write the authoritative version that everyone across the firm can use. Another important class of project is simply thinking through theoretical questions and taking complicated problems and trying to solve them in a sensible way.

What kind of research should i work on that’s interesting to jane street?

You should really work on whatever research that you find interesting and work on the problems that you want to solve. Most researchers come in with very, very little domain knowledge about prop trading and finance in general, and a lot of the learning happens on the job. So, a lot of the value in the research that you do before coming to Jane Street is really for your own edification and whatever you enjoy looking into. Just things you want to learn about the world. People had pretty varied interests before coming to Jane Street. In my case, I did some research on quantum information theory as well as machine learning for clinical applications.

What kind of classes do you recommend i take before interning at jane street?

There’s some background that’s useful to roughly every researcher in their day-to-day jobs. That would include some probability and some discrete math. Past that, it depends a lot on what you sort of consider your specialty and what your role might look like at the firm. In my case, I found optimization and some machine learning really valuable, and so I found those classes to be really helpful. In other people’s cases, that might include things like deep learning or theoretical stats or really a variety of other topics.

What type of machine learning does jane street do?

As you might imagine, I can’t really get into the details on the kinds of machine learning we do here. I will say that we generally try to stay on top of the state of the art. However, our problem domain tends to involve very low signal-to-noise ratios and really large amounts of data. And that means that we often develop techniques in-house that won’t really match what you see in the world at large.

How did you prepare for the interview process?

When preparing for a research interview, brushing up on probability and basic game theory will always be useful. We’re generally looking for candidates to demonstrate problem-solving skills. We aren’t looking for candidates to have domain knowledge of finance or prop trading. And so we generally will not test for those during interviews.