The RoleAs a Senior Data Scientist, own two things that make a regulated surveillance system credible: detecting information-driven trading, and proving the system actually works. The first is about timing - separating trading on public information from trading on non-public information, via the public-knowledge clock joined against pre-event positioning. On the US exchange this supports insider-trading detection at account grain; on the DeFi venue the same pipeline ports to wallet/cluster grain as behavioral pre-event positioning surveillance (deliberately without identity claims). The second is the assurance function a regulator-facing capability lives or dies on: rigorous, repeatable evidence that the detectors carry real signal rather than noise - now across two ground-truth regimes, including on-chain resolutions, which are public and deterministic. This is the most research-oriented of the seats and the closest to a quantitative-research profile. The second seat is junior and grows into the validation practice.
Key Responsibilities:- Build detection logic for insider and information-driven trading, centered on the timing of when information became public versus when it was acted on - on both venues, with the identity boundary each venue supports
- Design and run rigorous validation of the systems' outputs - statistical testing, permutation-based informativeness testing against market-resolution ground truth, backtesting against known cases - to demonstrate the detectors work
- Own the quality and trustworthiness of the labels that train the models, treating labeling as a continuously improving process rather than a fixed dataset - including the DeFi label corpus, which the program creates from zero
- Grow the validation work into a repeatable, audit-ready assurance capability as the systems expand to new markets
- Mentor more junior team members contributing to the validation and analysis work
RequirementsQualifications:- Strong empirical and statistical background: hypothesis testing, permutation / resampling methods, backtesting, and careful inference
- Experience working with weak, noisy, or evolving labels and human-in-the-loop labeling systems
- A quantitative finance or empirical-research background (including relevant PhD or equivalent industry experience) a strong plus
- Sequential-modeling experience; transformers or RNNs applied to behavioral or transaction sequences is a plus
- Production ML experience and the discipline that comes with regulated, audit-facing work
BenefitsPosition Location:This is an onsite position based out of our Santa Monica, CA or New York, NY offices.
Compensation: The base pay for this position is $190,000-290,000. A bonus will be provided as part of the compensation package, in addition to a full range of medical, financial, and/or other benefits.
