Jane Street does not run a standard bank interview. For trading and research it is a live reasoning conversation with probability, estimation, and odds that change mid-question. For engineering it is coding and systems depth. Here is what comes up and how to prepare.
Jane Street is a proprietary trading firm and liquidity provider, not an investment bank. Its interviews reflect that. The firm's public materials use trading games to show reasoning under changing information, and its engineering guidance points to coding and systems work rather than deal or valuation questions. So the first thing to get right is which conversation you are preparing for.
This guide covers how the process runs, the questions that come up by path, how to answer a probability prompt out loud, and a practice routine you can start this week. If you have not applied yet, start with our Jane Street internship application guide for paths, hubs, and timing, then come back here for the interviews.
How the Jane Street interview process usually runs
The stages vary by path and office, but most candidates move through the same shape. Confirm your own process, because some roles add a step.
- Online application. Resume and eligibility details, path and office selection, and on some listings a short set of questions rather than a full essay.
- First conversation. A recruiter or campus screen, then a first technical conversation with someone on the desk or team.
- Two to four technical rounds. These are live and interactive. Trading and research rounds focus on probability and reasoning; engineering rounds focus on coding and systems.
- Final rounds. Harder variants, more pushback, and a fit conversation with the team you would join.
The rhythm matters. Jane Street often keeps early conversations narrow: one probability thread or one coding problem, with follow-ups that go deeper rather than wider. Late rounds add interaction depth, and you may work through a problem alongside the interviewer rather than presenting a finished answer.
What Jane Street actually tests
Strip away the puzzle framing and every path is testing the same few things: whether you can reason cleanly under uncertainty, whether you can explain that reasoning, and whether you would be good to work with on a small team.
For trading and research, the core skills are probability, expected value, estimation, and mental arithmetic. You need to be comfortable saying "I would assume this, here is roughly what that gives me, and here is how I would check it." For software engineering and machine learning, the core skills are coding, data structures, and systems judgement, plus curiosity about markets.
Jane Street is known for using OCaml, but do not assume one language defines the whole interview. Prepare around the requirements in the live job description, and be ready to reason in whatever language the role names.
| Path | What the interview centres on | What it is not |
|---|---|---|
| Trading | Probability, expected value, estimation, calm updates | A deal walk-through or valuation case |
| Quantitative research | Structured reasoning, statistics, coding where listed | Generic "passion for markets" talk |
| Software engineering | Coding, systems, failure modes, trade-offs | Accounting or LBO mechanics |
| Machine learning | ML depth plus markets curiosity | Vague talk about AI in finance |
The questions that come up in trading and research interviews
Trading and research prompts usually start simple and get altered. Expect the family below rather than an exact list, because interviewers adapt to your answers.
- A game has a payout and a fee. What is the expected value, and should you play? What if the fee doubles?
- You draw items without replacement. What is the probability of a given outcome, and how does it change after one reveal?
- Estimate something with no data: how many of a product are sold in a city in a year?
- You are quoted a price on an event. Which side, if any, would you take, and why?
- A coin or dice game runs for a fixed number of rounds. What is the fair value, and where would you stop?
- Given a new piece of information, how should your estimate move?
The point is not to reach a perfect number. It is to show that you can set up a problem, name your assumptions, and update them cleanly when the interviewer changes a parameter. That is the daily work on a trading desk, so the interview is a small version of the job.
Our sales and trading interview questions guide covers the markets side of desk interviews, and the hedge fund interview questions guide covers stock pitch and strategy rounds if you are also applying to funds.
Questions for software engineering and machine learning paths
Engineering rounds look like strong software interviews at any serious technology firm, with a markets context. Expect coding, data structures, and system design with real constraints.
- Implement a function or data structure, then extend it when requirements change.
- How would you find a bug that only appears under load?
- Design a small system where latency matters. What would you trade off?
- Walk me through a project that failed. What did you learn?
- How would you test this, and what would you monitor in production?
Machine learning rounds add model depth: how you validate a model, how you spot overfitting or data leakage, and how you reason about a result that looks too good. Keep the answers grounded in projects you actually built, because interviewers probe the details quickly.
How to answer a probability question out loud
The biggest mistake is going silent to calculate. Interviewers cannot credit thinking they cannot follow. Use the same five steps on every reasoning prompt.
- Restate the question in one sentence so you and the interviewer agree on what is being asked.
- State assumptions before you calculate, especially for estimation prompts.
- Compute in small steps that the interviewer can interrupt or correct.
- Sanity-check the magnitude before you commit to the number.
- Invite pushback by naming how a change would move your answer.
Then hold that structure when the question changes. If the fee rises, say so and update. If the interviewer asks "are you sure?", re-check your largest assumption first rather than restating the whole calculation. Visible, correctable reasoning beats a silent answer that happens to be right.
| Interviewer move | Strong response |
|---|---|
| Changes a number mid-question | Name the change, update the calculation |
| Asks "are you sure?" | Re-check the biggest assumption first |
| Offers a shortcut | Adopt it and explain why it works |
| Pushes an edge case | State the boundary condition plainly |
A practice routine that works
Start early, because fluency comes from repetition rather than one long cram. Fifteen focused minutes a day beats an hour of silent puzzle grinding.
- Daily probability. Work one expected value or conditional probability problem and say the reasoning aloud.
- Estimation once a week. Take a Fermi-style question, state assumptions, and land on an order of magnitude.
- Markets conversation. Build one ninety-second view you can defend, including what would change your mind.
- Mock rounds. Once a week, work through problems with a friend who interrupts and changes the numbers.
- Coding, if engineering. Solve one problem under time pressure and explain your trade-offs out loud.
You can record the spoken parts in Finbound's Video Interview Prep to check pacing and clarity, though it is no substitute for live maths with another person. If you are also running bank processes, keep the prep separate: valuation practice is useful for investment banking interviews but it does not transfer to a trading round.
Mistakes that end a Jane Street interview
Most unsuccessful candidates lose the room in small ways rather than on one hard problem.
- Going silent to calculate instead of narrating the reasoning.
- Refusing to update when the interviewer changes the question.
- Bluffing a number instead of estimating with stated assumptions.
- Preparing investment banking valuation instead of path-specific skills.
- Treating the conversation as a test to beat rather than a problem to solve together.
- Ignoring the coding bar on an engineering application because the firm is "a trading firm".
Jane Street interviewers are usually trying to find out what it would be like to think through a hard problem with you for an afternoon. Curiosity, honesty about what you do not know, and a calm response to being wrong read better than a memorised script.
What to do after reading this
Preparing for a Jane Street interview? Pick the path from your live listing, then build a daily habit around the skills that path tests: probability and estimation for trading and research, coding and systems for engineering. Practise saying your reasoning aloud, and get comfortable changing your answer when the question changes.
Next reads: our Jane Street internship application guide, sales and trading interview questions guide, and D.E. Shaw internship guide.