A global insurance company has implemented a lightweight integer programming approach to optimize the assignment of online insurance policies to independent agencies. The system uses PuLP, an open-source Python library, to maximize productivity scores while balancing agency capacity, fairness, and geographic constraints. This replaces manual round-robin assignment methods that often led to delays and missed opportunities.

Historically, policy-to-agency assignments relied on sequential distribution with little consideration for capacity or expertise differences. The new optimization model translates agency performance into numeric productivity weights based on swap ratios, which capture value per unit of policy received. By incorporating constraints like capacity limits, ZIP code eligibility, and bucket-mix requirements, the system ensures assignments align with both operational capacity and business priorities.

The solution operates in batch mode for historical data analysis and real-time mode for immediate policy routing. Rather than waiting months for textbook-optimal models, the company chose a pragmatic approach that delivers deterministic, auditable results fast enough for real-time requirements. This modular system allows for future extensions with decomposition techniques or stronger solvers while maintaining its core contract. The approach demonstrates how basic optimization tools can solve high-value problems without requiring complex solutions.