Challenge

Co.opmart managed 100+ stores with 50,000+ SKUs relying on manual replenishment decisions. Stockouts caused $8M in annual lost sales while overstock tied up $15M in working capital, particularly for perishables.

Solution

ICTI built a lakehouse architecture on Databricks ingesting 5TB of daily sales, weather, event, and supplier data. ML demand forecasting models achieve 94% accuracy at SKU-store level, driving automated purchase orders and promotional planning.

  • Databricks lakehouse ingesting 5TB daily
  • ML demand forecasting — 94% SKU-store accuracy
  • Automated PO generation integrated with SAP
  • Real-time pricing engine for perishable optimization