End-to-end AI agents for property claims, from loss report to estimate. Backed by Y Combinator.
About the role
A claim is a hostile engineering environment. The data lives in Xactimate files, carrier portals, Zendesk queues, voicemails and photos of wet drywall. The workflows were designed twenty years ago and enforced by franchise agreements. SLAs are measured in minutes and breaches cost real money. Your job is to walk into that environment, reason from first principles about what the system should be, then build agents that run it in production.
What they're looking for
- Deploy into enterprise claims operations and own technical outcomes end to end
- Build production agent systems: prompt architectures, tool use, routing, evals, fallbacks, the glue between LLMs and systems that were never meant to have APIs
- Reverse engineer legacy claims software and make it programmable
- Design the data layer as you go. Every claim we touch generates hundreds of structured data points, and the pipelines you build become the moat
- Ship during live go-lives with hard dates and executives watching, including hurricane season surges where volume triples overnight
- Convert field solutions into platform primitives so each deployment is faster than the last
More about this role
Cozmo is the AI operating system for property claims. When a pipe bursts in someone's home at 2am, our agents answer the call, capture the loss, enter the claim into Xactimate and Cotality, dispatch the right contractor against SLA, chase acceptances before breach and draft the carrier-ready estimate from field photos. We run both halves of a claims operation: everything the customer touches and everything that happens behind the desk.
Our customers are restoration franchisors, TPAs and adjusting firms whose boards have told them to become AI-native and who have no way to do it themselves. Our anchor is one of the largest restoration franchisors in the US.
A claim is a hostile engineering environment. The data lives in Xactimate files, carrier portals, Zendesk queues, voicemails and photos of wet drywall. The workflows were designed twenty years ago and enforced by franchise agreements. SLAs are measured in minutes and breaches cost real money. Your job is to walk into that environment, reason from first principles about what the system should be, then build agents that run it in production.
You will spend part of your time inside customer operations, part of it writing code that...
Browse similar: AI jobs · AI startup jobs · Startup jobs