Hi there,
New to Retool and wondering if anyone has experience or advice on OCR?
In the flow I'm building I need to extract a few data points from a PDF document; in order to prefill some fields. Does anyone know if the following flow is possible or something similar?
- User uploads PDF via Retool app.
- PDF is sent to third party OCR tool (eg nanonets)
- Trained OCR tool extracts the data required from PDF
- Retool app receives the extracted data in assigned fields if successfully extracted and user can save and submit the data.
Is any of this possible with Retool or will I need to go with a custom internal solution?
Many thanks!
Hi @dominic-rg!
Welcome to Retool, I'd be happy to explore this you.
To confirm your flow, would something like this work for you?
- User uploads a PDF via your Retool app via the File Input component
- Make a POST request to the OCR tool's API endpoint, providing the PDF file as input.
- The OCR tool processes the PDF and extracts the required data points.
- The OCR tool creates a response with the extracted data in a structured format (e.g., JSON).
- In Retool, make a GET request to the OCR tool to get the extracted data
- Display the extracted data in Retool
To implement this, your OCR tool will need to offer an API integration and have endpoints that accept a PDF file via POST and can share data from a GET.
1 Like