As part of a consulting engagement with a major automotive manufacturer, I worked to develop and deploy two production-grade machine learning pipelines focused on vehicle delivery forecasting and automotive parts availability.
The first model estimates delivery timelines for vehicles arriving at dealerships based on production schedules, shipment data, and regional flow patterns. The second pipeline forecasts supply chain disruptions and part delivery delays using historical procurement, inventory, and logistics signals.
This work helped lay the foundation for scalable AI deployment across the client’s operational stack, and contributed directly to their internal enablement and MLOps readiness.
- Delivered two independently validated forecasting models for vehicle and parts prediction tasks
- Guided model architecture, training, and evaluation using real-world data pipelines and deployment constraints
- Partnered with client engineering and operations teams to define feature inputs, validate outputs, and support long-term maintainability
- Supported internal documentation and onboarding materials for continued AI integration
Tools & Stack
Python · SQL · Dataiku · Random Forest · XGBoost · MLOps Strategy · Forecasting Metrics
Demo and Access
Client-facing project. Demo and access of this project is propietary.