PDF to JSON, in your schema
Any PDF — invoices, contracts, bank statements, timesheets — returned as JSON that matches your exact field names, types, and computed fields. For pipelines feeding vector DBs or LLMs.
- You define every field name, type, and derived value in a JSON Schema — the parser respects them exactly
- Computes derived fields (totals, booleans, deduped lists)
- Same parser benchmarked on bank statements and invoices
- One API call: okra extract --schema schema.json
Related convert from pdf tools