Senior Data Scientist, Decision Science
MVF Global
In this role you will combine classical ML and decision science to optimize lead generation and traffic performance. You’ll work with AI engineers, PMs and operators to build, evaluate and deploy models that directly impact revenue. You’ll translate business rules into decision policies, learn from feedback, and set evidence-based guardrails before automation. This is a hands-on position within a small data science team focused on measurable commercial impact, experimentation and responsible AI deployment.
Pay / Benefits- Summer Fridays
- 25 days holiday + 8 bank holidays (increments to 30)
- Hybrid work - 3 days in office
- Work from anywhere for 2 weeks
- Life Assurance and Income Protection
- Health, dental, vision coverage allowance
- Build and productionize ML models for lead quality and traffic performance (classification, ranking, anomaly detection, forecasting).
- Evaluate, validate and monitor models offline and online (drift checks, calibration, leakage); collaborate with platform team for deployment on SageMaker.
- Quantify business impact via KPIs and A/B tests; establish measurement design and reporting.
- Frame business decisions as actions with context and reward; compare against existing rules; align with operators on optimization goals.
- Develop and evaluate policies using contextual bandits, safe exploration, and off-policy learning; incorporate operator overrides.
- Define evidence and guardrails for automatic decisioning; coordinate with business sign-off and engineers for enforcement.
- Collaborate with cross-functional teams to deliver robust AI-native solutions and scalable architecture.
- 5+ years as a data scientist or similar ML role with production models
- Strong statistics and experimentation experience (hypothesis testing, A/B testing, power analysis)
- Classical ML expertise: classification and ranking with scikit-learn, XGBoost or LightGBM; rigorous evaluation (validation design, calibration, leakage)
- Decision science experience: contextual bandits, causal inference or uplift modelling; ability to apply independently
- Programming and tooling: strong Python and SQL, Git and code reviews
- Candour and clear communication
- Delivery ownership and accountability
- Ability to work with cross-functional teams
- Python
- SQL
- Git
Reference: WJ-747_31213491 · Posted 01 Nov 2026, 12:16 PM