Accounts Receivable Operations
Led architecture and development for an AI-enabled collections application. Designed the core data model and knowledge graph relationships for three AI agents and one AI specialist supporting invoice prioritization, collection strategies, and payment reminders. The application was designed for 12 million invoices per year and 500,000 active business customers. Product management research projected about 30% less manual processing effort and 40% higher collector productivity. The application remained at the pilot stage.