Riddles in Supply Chain Management - How AI solves them?
summary
Shows an end-to-end system that uses demand forecasting with ML (feature engineering: holidays, weather, promotions, outlier and stockout handling), computer vision for cold-start SKUs, and evolutionary algorithms to optimize purchase orders, production sequencing and store transfers, deployed to Azure or on-prem and integrated back into ERP. Reports cutting manual work ~50% and lowering inventory 25–65% while reducing stockouts; safety stock is set dynamically from forecast error and BI dashboards explain and let users edit recommendations.
AI-written summary · about
chapters (19)
- 0:00AI in supply chain
- 1:34Supply chain challenges
- 3:19Traditional vs AI demand forecasting
- 5:44Machine learning for demand prediction
- 7:11Data for forecasting
- 9:17Forecasting riddles and solutions
- 11:02Seasonality and stockouts
- 12:05Promotions and outliers
- 13:09Price elasticity and pricing
- 14:12External impacts like COVID-19
- 14:55New items with computer vision
- 15:59Forecasts to purchase orders
- 17:02AI for production planning
- 18:05Store transfers and sales
- 19:09AI in ERP systems
- 19:30Financial impacts of AI
- 22:43AI industries and use cases
- 23:48Product overview and team
- 24:11Q&A on outliers and safety stock
related talks
27mLeveraging AI for Resilient Supply Chains
16mAutomate supply chains with agentic AI systems
27mManufacturing optimisation & planning using ML platform
16mTransforming Inventory Forecasting with Production ML
23mArtificial Intelligence for Store Assistance
16mAI Replays Decisions to Improve Supply Chain Performance
55mAutomate Inventory Rebalancing with Agentic AI
27mOptimizing Omnichannel Order Fulfillment with AI and Advanced Analytics