← Insurance
Statements of values and loss runs, mapped to your risk fields.
General document AI reads text. We map what it reads to how your underwriters rate risk.
InBroker documents in any layout.
OutStructured risk data with a confidence score on each field.
How it works today
- Statements of values and loss runs arrive in a different layout from every broker.
- Underwriters read them by hand to find the figures that matter.
- A keying error shows up at pricing, or later at claim.
What we build
- Step 1Reads statements of values, loss runs and schedules.
- Step 2Maps each figure to the risk fields your underwriters use.
- Step 3Scores every field for confidence.
- Step 4Queues low-confidence fields for a person to check, then writes the clean data to the file.
Where a person stays in charge
Anything below your confidence threshold is checked by a person before it reaches the file. You set the threshold.
Connects to
Your underwriting file, rating inputs and document store.
What we measure
- Share of fields needing correction
- Time to extract per submission
- Size of the review queue
- How many underwriters use it each week