Reinforcement learning optimizing working capital and freeing $28 million
A $4 billion retailer carried 92 days of inventory and $180 million in excess cash.
StartDate deployed a reinforcement-learning agent that simulates thousands of working-capital scenarios daily, optimizing payment terms, inventory levels, and dynamic discounting. The system auto-executes approved actions via API.
Within 12 months:
• Freed $28 million in trapped cash
• Reduced DIO from 92 to 61 days
• Generated $11 million in new early-payment discounts