What interns at Jane Street actually do, from 2016 to now

The best path to a lucrative job at Jane Street is also the most obvious: get an internship and convert it. But what do Jane Street interns actually do? The trading firm has made a habit of shouting out some of its most impactful interns within its engineering teams each year since 2016 in a yearly blog post. These have given us multiple insights into both how the internships work and what kinds of projects you'll find yourself working on.

💥Follow us on WhatsApp for news alerts.💥

How do Jane Street internships work?

The broad structure of Jane Street engineering internships has remained constant. Developers stay with the firm for three months and are given two projects to complete, one for each half of the internship.  A separate blog post from Jane Street in 2022 said that the firm asks full-time employees to come up with these projects "based on the needs of their team." Interns then have an in-depth conversation with the internship organizers to understand their strengths and interests; they are assigned projects and mentors which are the "best possible fit for them." The mentor assigned is also the person who originated the project.

These two projects can sometimes be very different. One intern in Jane Street's 2016 blog had two very different projects: a UI tool used by traders and a set of OCaml extensions used by engineers. 

Jane Street has branches of engineering which run their own separate internship programs, such as machine learning research and engineering or Linux engineering and IT. These also operate using a project-based system.

This year, Jane Street also shed some light on what its trading desk operations engineers (TDOEs) interns do. The nature of the role is different - TDOEs work on building more efficient workflows for trading teams, but are also on call to handle "the urgent needs of trading." The structure for these internships is different; Jane Street said they complete three separate projects, then spend two weeks "sitting with teams." These internships also often give you access to "sensitive details about [Jane Street's] trading systems or strategies."

Jane Street's strategy and product (SP) team also operates differently. SP interns are "embedded directly onto a team, paired with a full-time mentor, and given real-world projects we actually care about getting done." SP internships require "a unique combination of big-picture thinking... and deep analytical problem-solving."

What do interns actually work on? 

Across 11 total blog posts, Jane Street has mentioned 80 different intern projects (26 of which were in this year's edition). While heavily curated, they give a taste of what kinds of projects are most common.

The most commonly mentioned type of projects involved developing new tools for the firm's programming languages. This has historically meant building new functionality for OCaml, the niche language Jane Street champions, often borrowing features from other popular languages like Rust. The firm has also rapidly adopted Python because of AI; one intern this year focused on improving the RAM efficiency of PyTorch, a python library used in machine learning. 

AI projects were the next, most common. You could also find yourself working on Jane Street's data infrastructure, or on its networking capabilities. 

How have Jane Street internships changed?

As you would expect, AI has played a major role in Jane Street internships, both in terms of training and the projects themselves. While AI projects are nothing new, the firm recently added a machine learning engineering internship program alongside its machine learning research track. 

ML engineering projects are more in line with traditional engineering ones whereas research projects are "more exploratory—they are, genuinely, research." One engineering intern "developed an activation checkpointing scheme... in order to keep model training runs within a tight RAM budget." A research intern, by contrast, experimented with a variety of ways to stop LLMs from memorizing results during pretraining to make backtesting more accurate; that intern used a variety of methods including "divergence decoding, distillation, natural language autoencoders, and some prompt engineering."

More broadly nature of intern projects has become more ambiguous because of AI coding tools. The 2022 blog said that Jane Street previously focused on projects where the intern "produces code early" so they could get more immediate feedback; Developer educator Aaron Bauer said on the firm's podcast last month that some of these projects could now be solved with a single prompt. Instead, Bauer said Jane Street is focused on providing  "more open-ended" projects. One 2026 intern working on an indexing strategy for Jane Street's messaging system "prompted an agent to spin up five different [implementations of his project] and profile them overnight." He also "had a swarm of agents analyze the code with high scrutiny."

The structure of training has also changed to place increased focus on review[ing] code as well as writing it. Bauer said interns now have "three additional days of training in the middle of the internship on code review," where they analyse "pre-made features actually adding functionality to the same application that they were doing the testing and design work on." Interns also conduct code reviews on the projects of other interns, and mentors will send them additional pieces of code to review.  Bauer said that Jane Street is also using the time saved on coding to train interns on "topics that maybe didn’t get a lot of formal training at the start of the internship, like testing."

Have a confidential story, tip, or comment you’d like to share? Contact: WhatsApp: http://wa.me/442079977910 (+44 20 7997 7910), Telegram: @AlexMcMurray, Signal: @AlexMcMurrayEFC.88 Click here to fill in our anonymous form, or email editortips@efinancialcareers.com.

Bear with us if you leave a comment at the bottom of this article: comments are moderated intermittently by human beings. Sometimes these humans might be asleep, or away from their desks, so it may take a while for your comment to appear. You must take sole responsibility for comments you post on this site. We will take reasonable steps to weed out anything that we consider to be offensive or inappropriate.