Supply Chain Data Scientist (Optimization & Forecasting)

Senior Supply Chain Data Scientist (On-Site)

Location: Analytics Center / Office (Full-Time, 100% In-Office)

Experience: 4–8 Years

Department: Advanced Analytics & Operations Research

About the Role:

We are looking for a quantitative Supply Chain Data Scientist to build the mathematical engines that power our predictive logistics, network optimization, and automated inventory management platforms. Working hand-in-hand with our data engineers, software architects, and client success leads at our central office, you will transform massive, multi-enterprise logistics datasets into actionable, real-time decision frameworks that help global enterprises minimize stock-outs, reduce holding costs, and bulletproof their supply chains against disruption.

Key Responsibilities & Detailed Breakdown:1. Demand Forecasting & Predictive Modeling

• Design, train, and deploy advanced machine learning models (gradient boosting, deep learning time-series architectures, and probabilistic estimators) for multi-horizon demand forecasting and SKU-level volatility analysis.

• Build predictive models to anticipate supplier lead-time fluctuations, customs clearance bottlenecks, and transportation carrier delays before they impact downstream production schedules.

• Continuously monitor model performance, analyze feature drift, and retrain pipelines using automated data feeds from enterprise ERP and WMS integrations.

2. Operations Research & Network Optimization

• Formulate and solve complex operations research (OR) optimization problems, including multi-echelon inventory optimization, safety stock placement, and network footprint design.

• Develop mathematical algorithms for dynamic vehicle routing, last-mile delivery scheduling, and freight consolidation using optimization solvers (e.g., Gurobi, PuLP, or SciPy optimization libraries).

• Build "digital twin" simulation models of physical supply chain networks to test demand shocks, supplier failures, and inventory rebalancing strategies virtually before production deployment.

3. Productionizing Analytics & Feature Engineering

• Collaborate closely with Data Engineers to design efficient feature stores and data transformation pipelines in Python and SQL that feed high-frequency scoring engines.

• Translate experimental Jupyter notebook models into clean, modular, and containerized microservices ready for enterprise-grade deployment.

• Optimize algorithmic query execution and vector/matrix calculations to ensure sub-second response times for interactive user dashboards.

4. Cross-Functional Collaboration & Client Workshops

• Engage in daily face-to-face whiteboarding and sprint planning sessions with software engineering squads to align model outputs with core product features.

• Partner with product managers and client-facing teams to interpret analytical findings, validate assumptions against real-world logistics constraints, and present optimization insights directly to enterprise clients during on-site consultations.

Required Skills & Qualifications:

• Work Mode: 100% On-site commitment with daily physical attendance at our office and analytics center.

• Core Technical Stack:

  • Expert-level Python (Pandas, NumPy, Scikit-Learn, PyTorch/TensorFlow, Statsmodels).

  • Strong mastery of relational databases and complex SQL querying.

  • Experience with Operations Research solvers (Gurobi, CPLEX, or SciPy optimization).

  • Familiarity with containerization tools (Docker) and version control (Git).

• Domain Expertise: Deep practical knowledge of core supply chain metrics and concepts (Fill Rate, Days Sales of Inventory, Inventory Turns, Bullwhip Effect, and multi-echelon inventory routing).

• Education & Background: Advanced degree (Master’s or Ph.D.) in Operations Research, Industrial Engineering, Data Science, Applied Mathematics, or a related quantitative field.

Why Join Us On-Site?

Collaborating side-by-side with domain experts and software teams in a physical office setting enables rapid mathematical whiteboarding, immediate feedback loops on model constraints, and direct participation in high-stakes enterprise analytics implementations.