Define the STP eligibility criteria as declarative rules: acceptable SIC/NAICS codes, revenue thresholds, building age limits, loss history requirements, and geographic restrictions
Load the rules into a business rules engine (e.g., Drools, Camunda DMN, or a vendor underwriting workbench) and version-control the rule set
At submission time, call the rules engine with the application data and capture the decision output: approve-STP, refer-to-underwriter, or decline
For STP-approved submissions, automatically trigger quote generation, document issuance, and policy binding without manual intervention
For referrals, create an underwriter task with the specific rule that triggered the referral and the relevant risk attributes
Track STP rate (percentage of submissions that bind without referral) as a key metric and use it to calibrate rule thresholds over time
Known gotchas
Rules that look at external data sources (MVR, property, prior loss) introduce latency and failure modes; design the STP flow to handle external data unavailability gracefully without blocking all submissions
Underwriting rules are subject to rate filing and form filing requirements in many states; changes to eligibility criteria may require regulatory approval before deployment
Over-aggressive STP rules can increase adverse selection; monitor loss ratios on STP-bound business separately from referred business to detect drift
Give your agent this knowledge — and 200+ more routes
One MCP install gives any agent live access to the full route map, with trust scores updated by agent consensus:
claude mcp add --transport http waymark https://mcp.waymark.network/mcp