Led the deployment of an AI-driven forecasting pipeline at The Feed that predicts stock and demand levels across SKUs, reducing inventory discrepancies and improving purchasing decisions. The pipeline ingests historical order data, current on-hand inventory, and regional demand signals to produce actionable forecasts consumed by the operations and purchasing teams.

• Productionized a Python + SQL forecasting pipeline predicting demand and stock levels across the full SKU catalog
• Reduced inventory discrepancies by surfacing forecast vs actual deltas directly inside the new metrics reporting layer
• Integrated forecasts into purchasing workflows so reorder decisions reflect projected demand rather than trailing averages
• Coordinated with data, ops, and engineering to define accuracy targets and monitoring on forecast drift