Use cases¶
Each page has a complete, working example.
| Extract data from invoices | Turn PDF and scanned invoices into typed data (vendor, invoice number, dates, line items, totals) with a citation for every value, in Python, using Claude Haiku or any model. |
| Extract transactions from bank statements | Parse bank statement PDFs into a list of transactions (date, description, amount, balance) with citations, using an open-source Reducto alternative. |
| Extract data from receipts | Read photos and scans of receipts into merchant, date, items, tax and total with citations, using Claude Haiku and an open-source parser. |
| Extract data from tax forms (W-2, W-9, 1099) | Pull fields and checkboxes from US tax forms such as the W-2, W-9 and 1099, scanned or fillable, with word-level citations. |
| Extract tables from PDFs | Convert tables in PDFs and scans to HTML, Markdown or structured rows, including merged cells and tables that span pages, with Python. |
| Convert PDF to Markdown for RAG and LLMs | Turn PDFs, scans and Office files into clean Markdown with headings, lists and tables, ready for chunking, RAG and LLM context. |
| OCR for scanned documents and handwriting | Read scanned PDFs, faxes, old typewritten pages and handwriting into text and data with a vision model plus local OCR checks. |
| Extract clauses and terms from contracts | Pull parties, dates, terms and clauses out of long contracts and agreements (PDF or Word) into structured data with citations. |
See them on real pages: Real documents, end to end.
If ReadAnyDoc saved you time, a star on GitHub helps other developers find it.