Policies and extraction
Upload a policy and BindCheck reads the whole PDF — the declarations, the schedule of forms and endorsements, and the coverage details behind them. Here's what happens and how to intervene when a document is unusual.
What gets extracted
- Line of business, carrier, named insured(s), policy number and effective/expiration dates.
- The form schedule — every ISO/AAIS form and endorsement by number and edition date, with a short plain-English description of purpose. (We store form FACTS only; copyrighted form wording is never reproduced.)
- Declarations fields — limits, sublimits, deductibles, retentions, coinsurance and covered-auto symbols as applicable to the line.
- Additional insureds and key endorsements (waiver of subrogation, primary & non-contributory) identified by form.
- An extraction-confidence signal; low-confidence or unusual documents are surfaced for your review rather than guessed.
Processing, retries and errors
Extraction runs in the background — uploads return instantly and the page updates itself. Failed reads retry automatically. A policy stuck in error usually means a corrupted or image-only scan: re-run extraction on the policy, or fix the source PDF and upload again.
Correcting a read
Open the policy → Edit extraction to correct the named insured, dates, a limit or deductible, or a form entry. Corrections re-run the check immediately and are recorded in the activity log. Extraction is accurate on clean policy PDFs; corrections mostly matter for unusual or partly-illegible scans.
Manuscript and non-standard forms
When a form doesn't map to the standard library it's captured as a manuscript endorsement and flagged for human review — the engine never pretends to interpret bespoke wording. You read the language against the account's needs.