Automating Invoice Processing with OCR: Issues Extracting Data Using OpenAI API

Hello!

I’ve been pondering how to create an automation to process my invoices using OCR.

The invoices could be physical receipts that I’ll photograph with my mobile phone, send them to a Telegram chat, and this will trigger an automation that should process the data from the photo, create a JSON (date, amount, concept…), and record the data in a Google Sheet.

The problem I’m encountering is in extracting data from the image using OpenAI’s API. The response I’m getting says it can’t help me with that when I send the photo (vision module).

Has anyone tried something similar or know how I can do it?

Thank you in advance.