Ingest raw telematics trip data from a connected car API, mobile SDK, or OBD dongle: timestamp, GPS coordinates, speed, acceleration, braking, and cornering events
Calculate per-trip behavioral scores using a configurable scoring model: hard braking frequency, rapid acceleration, speeding percentage, distracted driving indicators, and nighttime driving proportion
Aggregate trip scores into a policy-level UBI score over the scoring period (typically 90 days for initial discount, then ongoing for renewal)
Map the UBI score to a discount tier per the filed rate plan and apply the discount to the renewal premium
Store the scored trip data with appropriate retention controls and provide an insured-facing score dashboard showing driving behavior breakdown
Handle opt-out requests by reverting to the base rate and purging telematics data per the privacy policy
Known gotchas
Telematics scoring models must be filed with state DOIs as part of the rating plan; using an undocumented scoring algorithm or applying a discount that was not filed can constitute unfair discrimination under insurance rating law
GPS data gaps and phone-as-sensor limitations produce false hard-braking or speeding events; implement a noise-reduction filter and define a minimum data quality threshold before scoring a trip
Privacy regulations in some states impose strict requirements on telematics data retention, purpose limitation, and disclosure; confirm compliance with applicable state telematics privacy laws before launch
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