Insurance Forms: From PDF Fields to Nested JSON

The challenge

Turn two insurance forms into records a claims system could use. Start with the model answer, then try the second form yourself. Keep the status and incident type exactly as printed.

Start here

Open the SharePoint folder linked above. Model_Insurance Claim - CLM-O6W4MRE71.pdf has the example answer claim_CLM-O6W4MRE71.json. Insurance Claim - CLM-000001.pdf is a different claim to process yourself. Both PDFs have two pages and selectable text. Inspect both pages before choosing OCR.

Try it

1. Extract the model form into the included JSON schema, including every expense row. Identify printed fields that the model JSON omits.

2. Apply the mapping to CLM-000001 and produce its own JSON record.

3. Verify both claim totals and calculate each reporting delay in days.

4. Find a contradiction between a structured field and the narrative without silently rewriting either value.

What to build

Build two claim records, an expenses CSV keyed by claim_id and expense_index, and a short validation report that separates extraction mistakes from inconsistencies already present in the source.