12 days ago
London, United KingdomStaff+
Responsibilities
- Own the real-time transaction decisioning system and its underlying machine learning platform.
- Develop feature infrastructure for batch, near-real-time, and in-request processing with defined freshness budgets.
- Maintain alignment between offline training and production serving, and build feedback loops capturing decisions and outcomes.
- Operate low-latency transaction-scoring services with robust fallback and degraded paths.
- Improve replay, shadow, and staged-rollout tooling so live model changes are safe and reversible.
- Own models from training through retirement and detect model decay before it causes financial losses.
- Lead through ambiguity, establish engineering standards, conduct rigorous reviews, and mentor engineers.
Requirements
- Experience building and operating high-availability services on critical paths within strict latency budgets, including robust fallback mechanisms.
- Strong systems-thinking skills covering failure points, graceful degradation, dependency failures, and feedback loops.
- Ability to write tested, typed, maintainable code that handles duplicate, late, and out-of-order events.
- End-to-end ownership of production feature or data pipelines, including resolving discrepancies between offline and production metrics.
- Experience turning ambiguous problems into shipped work and increasing the effectiveness of other engineers.
- Preferred experience with decision explainability, audit trails, attribution, LLM-driven analyses, or communicating model behavior to non-technical audiences.
- Preferred experience developing anomaly detection systems for novel attack patterns and emerging abuse without existing labels.
- Familiarity with GCP, BigQuery, Bigtable, Memorystore, Vertex AI, and Kubernetes.
Benefits
- Competitive salary package and equity ownership, including a performance-based equity bonus.
- Pension contributions from day one, private healthcare, enhanced parental leave, birthday leave, flexible time off, and Wellhub wellness membership.
- Hybrid working with approximately 1–2 office days per week in London, with fully remote work also supported; relocation is not available.
- Commuter benefits, lunch credit on office days, home office setup allowance, and a remote working allowance.
- Unlimited enterprise access to Claude, ChatGPT, Gemini, and other AI tools.
- Monthly product budget, zero-fee crypto transactions, and a $1,000 annual training budget.
- Employee referral program, High Potential Program, regular remote company offsites, and applicable UK/Ireland commuter and EV benefits.
Tech Stack
Categories
