How I built the Invoice Processing demo

The problem

Invoices arrive as PDFs, scanned images saved as PDFs, Word docs, or plain email text pasted into a form. Most invoice-AI demos quietly assume the first case. Real AP inboxes get all of them, plus the occasional file that isn't actually an invoice at all. This demo currently covers PDFs, DOCX, and plain text with a real text layer — a scanned image with no embedded text (a photo, or a PDF that's really just a picture) needs an OCR step this pipeline doesn't run yet, and isn't demonstrated here.

The pipeline

File upload hits a shared /api/extract-text route before the invoice-specific logic ever runs:

1. 01Detect file type by extension, not MIME type — browsers report MIME types inconsistently across OS/browser combinations, extensions don't.
2. 02PDF → unpdf (serverless-compatible; pdf-parse and pdfjs-dist both hit Node API assumptions that break under Vercel's runtime — this took two failed attempts to learn).
3. 03DOCX → mammoth, extracting raw text only.
4. 04TXT/MD/CSV → read directly, no parsing needed.
5. 05Text truncated to a 10,000-character ceiling before it ever reaches a model call.

The extraction call

The prompt asks for one JSON object — vendor, invoice number, dates, PO number, line items, subtotal, tax, total, payment instructions, and a self-reported confidence score — and nothing else. The response is regex-matched for the outermost{...}block before JSON.parse, since models occasionally wrap output in a sentence even when told not to.