Data Scientist, London

About the Role

We're hiring a mid-level Data Scientist (2 to 5 years' experience) for our Data team, working within Risk & Fraud. The brief is simple: reduce fraud without adding friction for good users. In practice that means ML models, deterministic rules and signal tuning, and working directly with our external risk vendors. You own that work end to end, from the question, to what ships, to the decision leadership makes off the back of it. Risk and fraud is where you'll have the clearest impact, but the role reaches across Product, Growth, and Engineering, and your work turns into product, policy, and revenue. If you're drawn to crypto, payments, and the kind of data they throw off, there's a lot here to get into.

What You'll Do

• Risk, fraud and compliance: Build and iterate on fraud detection, chargeback prediction, and transaction-risk models. Develop features and rule sets that work alongside our Risk and Compliance teams to keep bad actors out without adding friction for good users.
• Product analytics and growth: Own funnel analytics across on-ramp and off-ramp flows. Design and analyze A/B and multivariate experiments, identify conversion bottlenecks (KYC, payment method, geo), and partner with PMs and designers to ship measurable improvements.
• ML/AI modeling: Design, train, and deploy machine learning models, from classification and forecasting to clustering and recommendation, that power decisions inside the product (e.g., dynamic payment method ranking, user lifetime value, churn prediction).
• Business intelligence and reporting: Build trusted dashboards and self-serve data products for Product, Growth, Finance, and the executive team. Define and steward the metrics that the business runs on.
• Storytelling and strategy: Turn analyses into clear narratives and recommendations. Present findings to engineers, PMs, and the C-suite alike, and influence roadmaps with data.
• Data craftsmanship: Partner with Data Engineering to improve event tracking, data models, and the warehouse. Treat data quality as a first-class product.

Requirements:

• 2 to 5 years of experience as a data scientist, analytics engineer, or quantitative analyst, ideally at a fintech, payments, marketplace, or consumer tech company.
• Strong SQL. You can navigate large, messy warehouses (BigQuery, Snowflake, Redshift, or similar) and write performant, readable queries.
• Solid Python (or R) for analysis and modeling: pandas, scikit-learn, statsmodels, and at least one deep-learning or gradient-boosting framework (XGBoost, LightGBM, PyTorch, TensorFlow).
• Experimentation fluency. You understand the math behind A/B testing, sample sizing, power, and common pitfalls (peeking, multiple comparisons, novelty effects).
• Machine learning intuition. You can pick the right model for the problem, evaluate it honestly (precision/recall trade-offs, calibration, drift), and ship it responsibly.
• Visualization and BI. Comfortable building dashboards in Looker, Metabase, Tableau, Superset, or similar.
• Communication. You can explain a confusion matrix to a PM and a funnel drop-off to the CEO, in the same week, in the same tone.
• Ownership. You treat ambiguous problems as opportunities and don't wait to be told what to analyze next.

Nice to Have

• Experience in crypto, payments, banking, fraud, or compliance.
• Familiarity with dbt, Airflow, or similar data-stack tooling.
• Exposure to causal inference (difference-in-differences, propensity scoring, uplift modeling).
• Experience deploying models to production (batch or real-time) alongside engineers.
• Knowledge of AML / KYC frameworks or experience working with regulators.