The challenge
First notice of loss arrived as photos, PDFs and LINE messages. Adjusters re-typed the same facts into the core system, and simple claims waited behind complex ones with no consistent triage rules.
What we did
Embedded a 7-month AI Implementation Team to build FNOL capture, document classification and a triage scoreboard for adjusters. Added a project-based claims-AI specialist for four months to harden prompts against policy wording and train supervisors on exception queues.
How it ran
- 1
Discovery on two claim types
- 2
Pilot with dual-run (AI suggest + human decide) for six weeks
- 3
Scale after accuracy gates
Risks handled
- PII redaction before any cloud model call
- Change champions in each branch
- Exportable prompt packs so the insurer can swap vendors later
Want results like these?
Tell us the workflow that costs the most. Within two business days you get a recommended hiring model and a scoped proposal.
Talk to us