Jane Street

Jane Street describes itself as a global liquidity provider and trading firm using quantitative analysis and market mechanics to support consistent prices.

The firm says it trades continuously on more than 200 electronic exchanges and other venues; this is a company assertion, not an independently audited count.

The employer page reports more than 3,000 staff across five global offices; this is a firm-wide figure, not a London team size.

The London internship vacancy is labelled Trading, Research, and Machine Learning, with Quantitative Trading as the team.

A historical Jane Street technical article describes a machine-learning seminar in its London office. This establishes a London technical example, not the current intern team's stack.

Jane Street frames research, technology and trading as intentionally overlapping in its London quantitative trading internship.

Jane Street's quantitative research page says its team mostly writes in OCaml; this is a work-method distinction, not a claim about every London intern.

The public Jane Street Core repository is an open-source OCaml standard-library overlay, but its existence does not prove any particular internal deployment or intern assignment.

Quantitative researchers describe analysing large datasets, testing models, creating trading strategies and writing the code that implements them.

The ML team describes building neural-network models for trading and infrastructure for model training and inference.

A company technical article says Jane Street has revisited formal methods after a long period of scepticism; this reflects an engineering investigation, not a deployed firm-wide guarantee.

A published Jane Street ML example visualises how ReLU networks divide a plane into piecewise-linear regions; the page does not claim this produces a trading signal.

A historical London seminar reproduced classical deep-learning examples with functional OCaml models trained on a GPU through TensorFlow; it is a dated experiment, not evidence of today's production architecture.

Jane Street's public ML material presents deep learning as a priority for quantitative trading, while giving no independently verified impact on trading returns.

The formal-methods article indicates exploration of stronger software verification techniques, but provides no dated implementation milestone for London trading systems.

Jane Street itself highlights edge cases and tail risk when markets deviate from expectations, so a market-making description alone should not imply low-risk earnings.

The ML team calls financial data mostly noise, which makes apparent predictive patterns particularly vulnerable to overfitting.

Jane Street's historical OCaml ML experiment notes rough edges in TensorFlow's OCaml bindings; it should not be read as proof of today's production ML architecture.

The current native vacancy is explicitly a London Quantitative Trader Internship in the Quantitative Trading team.

The role says interns are paired with experienced traders to learn market signals, strategy execution, modelling, statistics and trading intuition.

Examples in the vacancy include datasets, predictive models and strategy simulations; the page presents these as past projects rather than guaranteed assignments.

The general internship page lists several disciplines across teams and offices; those programme-wide offerings do not establish a specific London placement outside the native role.

The native London vacancy heading specifies a June–September internship period but no cohort year; the period should not be converted into a verified 2027 intake.