Project detail
Inventory and Logistics with PyMC
Bayesian model for demand, inventory, and logistics lead times focused on reducing stockouts and costs.
Highlights
- Inventory optimization with probabilistic demand
- Applied to routes and delivery times
- Better planning under uncertainty
- Reduced stockouts and overstock
No public repository or demo link is published for this project yet.
Architecture
Applied guide with Bayesian models for historical demand, optimal inventory estimation, and support for logistics operations
Challenges
- Capturing demand variation without making the model too complex
- Translating predictions into actionable stock levels
- Incorporating external factors such as traffic or weather
Outcome
A foundation for better logistics decisions with less reliance on fixed averages
Overview
Problem
Logistics suffers when inventory and deliveries are planned with rigid assumptions instead of explicit uncertainty.
Approach
I used PyMC to estimate demand and derive inventory levels that are more robust to real-world variability.
Outcome
The work helps reduce stockouts and plan operations with more safety margin.