Extract Thai PDFs to Google Sheets with Typhoon OCR & Llama 3.1

This workflow processes PDF or scanned documents in Thai language by leveraging Typhoon OCR, one of the most accurate tools for Thai text recognition. It starts with a manual trigger or automation source, reads files from a local 'doc/' folder, executes Typhoon OCR via Python commands to generate markdown output, and then uses Llama 3.1 (or compatible LLM like GPT-4) to intelligently extract structured fields such as names, dates, and other key data.

A code node parses the AI's JSON-formatted response, ensuring clean data handling, before appending rows to a Google Sheets spreadsheet. This end-to-end automation eliminates manual data entry, reduces errors from poor OCR on Thai scripts, and integrates seamlessly into enterprise pipelines.

Benefits include massive time savings for HR, government, or operations teams handling Thai paperwork, high accuracy for complex layouts, and scalability for batch processing. Use cases: local government forms, enterprise HR onboarding docs, invoice processing in Thai businesses, or any AI-driven document automation requiring Thai support. Requires self-hosted n8n with Typhoon OCR installed.