Your new company
Havi, a global leader since 1974, employs over 10,000 people and serves customers in more than 100 countries. Specializing in the food service industry, Havi provides innovative supply chain and logistics solutions, including analytics, planning, distribution, and freight management.
Havi Supply Chain Tech Hub’s teams collaborate seamlessly across locations and functions, embodying a spirit of integrity and creativity to serve their customers in the best way possible.
Your new roleAs a Data Scientist, you will play a key role in shaping data-driven decision-making across the supply chain. You’ll design and operationalise predictive and optimisation models that directly impact areas such as demand and ETA forecasting, exception prioritisation, cost-to-serve and inventory optimisation. Working from certified domain data products, you’ll collaborate closely with the Security, Intelligent Automation & AI COE team to deploy models in batch, real-time and RAG-assisted workflows.
Your insights will be embedded into Intelligent Automation pods and BI tools, driving measurable value and enabling smarter, faster operations.
You will be responsible for:
Problem Framing & Prioritisation
- Translate domain OKRs into modelable problems with target metrics, baselines and success thresholds
- Estimate value and effort in collaboration with the triad
Data Discovery & Feature Engineering- Pull from certified data products and semantic layers
- Create robust features and document assumptions, leakage risks and data contracts with Analytics Engineering
Modelling & Evaluation
- Build and compare models (forecasting, classification, ranking, optimisation)
- Apply proper cross-validation and backtesting; define offline metrics and guard against bias and leakage
Experimentation & Decision Science
- Design and run A/B or alternative experiments
- Quantify impact (uplift, cost/benefit) and partner with PMO to capture benefits
Operationalisation (with AI Platform & Safety)
- Package models for batch/stream serving with defined SLAs, inputs/outputs and monitoring hooks (quality, latency, cost)
- For LLM/RAG-assisted flows, design retrieval strategies and evaluation harnesses; apply guardrails per policy
Monitoring, Drift & Lifecycle
- Set up dashboards and alerts for data/feature/model drift, performance decay and incidents
- Plan retraining, versioning and maintain model cards and deprecation plans
Governance, Privacy & Risk
- Produce model risk documentation (intended use, limitations, safety checks)
- Ensure approvals for sensitive data and compliance with access, retention and audit policies
Adoption & Automation
- Embed model outputs into Intelligent Automation workflows, copilots or BI tools
- Create playbooks and user guidance; iterate based on feedback
Collaboration & Enablement
- Partner with Analytics Engineering to productionise features in the warehouse
- Work with Governance to close data quality gaps and contribute reusable components and templates
What you will need to succeed
- Bachelor’s in Data Science, Statistics, Applied Mathematics, Computer Science, Operations Research or related field
- Master’s in Data Science, Statistics, OR, Analytics (or equivalent) is a plus
- Certifications such as Cloud ML (AWS, GCP, Azure), MLflow/MLOps, experimentation/causal inference coursework; dbt Fundamentals (optional)
- 4–8+ years of applied experience delivering data science models used in production with measurable business impact
- Proficient in Python (pandas, numpy, scikit-learn), plus experience with at least one of XGBoost, LightGBM or PyTorch
- Hands-on experience with feature engineering on warehouse/lakehouse data (Snowflake, BigQuery, Databricks), in close collaboration with Analytics Engineering
- Comfortable with experimentation methods (A/B testing, quasi-experiments), model evaluation (regression, forecast, classification metrics), and model risk documentation
- Familiarity with LLM/RAG evaluation and safety principles (using market LLMs; no custom LLM training)
- Exposure to supply-chain domains such as orders, shipments, lanes, OTIF, dwell time or inventory is a plus
- Strong communication, collaboration and delivery mindset: challenges the status quo, adapts quickly, promotes teamwork, anticipates risks and drives outcomes
- Clear storytelling with data, including concise model cards, decision memos and impactful visuals
- Solid software hygiene: version control (git), code reviews, reproducible environments and testing practices
- Experience writing inference-ready code with awareness of production constraints (latency, throughput, cost, drift)
- Privacy-aware modelling practices, including minimising PII, applying masking/pseudonymisation and respecting retention policies — especially in HR or Legal contexts
- English Fluency is mandatory.
What the company can offer you
Have the opportunity to join a cross-functional team in an international company with a multicultural working environment!
Next Steps
If you are interested in this opportunity, please send us your updated CV. If you are looking for another type of professional challenge, please contact us to discuss other career opportunities, always in complete confidence.