Skip to main content
← Back to all work
Document Processing

Universal Document Text Extractor

Document-to-text workflow built with n8n and LlamaParse. Upload a PDF, Word file or scan and get clean, structured text back, ready for AI agents and downstream automation.

Project Media
1 / 2

The Problem

Documents arrive in mixed formats (PDFs, Word files, scans), and getting usable text out of them is slow and inconsistent. Copy-paste breaks layout, scans return nothing, and every format needs its own handling. That makes the text hard to feed into AI agents or automations, which need clean, predictable input.

What I Built

I built an n8n workflow around LlamaParse. A user uploads a document through a form and the file is sent to LlamaParse for parsing. The workflow checks the job status and waits until parsing completes, then fetches the clean text. The output is structured and ready to pass to an AI agent or the next automation step.

The Result

Working document-to-text pipeline, tested on Word documents. It turns an upload into clean text ready for AI agents, with no manual copy-paste.

Stack
n8nLlamaParseDocument AutomationREST APIJavaScript
More Projects
Document Processing

Engineering Drawing Material Extraction Workflow