Automated Knowledge Extraction via Google Drive and InfraNodus

This n8n workflow streamlines the process of extracting, categorizing, and analyzing content from PDFs, Markdown, and text files stored in Google Drive. The primary goal is to enable a seamless interaction with document content through InfraNodus’ knowledge graph platform, bypassing traditional vector databases.

The workflow begins with searching a specific Google Drive folder for relevant files, then iterates over each file, downloading and identifying its MIME type. Based on the file type, it extracts the content using dedicated nodes for PDFs, Markdown, or plain text files. The extracted data is subsequently mapped and sent to InfraNodus’ API to visualize and build a knowledge graph representation of the document.

Additional steps include optional conversion of PDFs to higher-quality text using ConvertAPI, and a manual trigger for testing document ingestion. The process also facilitates a conversational interface, allowing users to ask questions about the content or generate prompts through AI nodes connected to OpenAI and LangChain, integrated with InfraNodus to provide contextual, topically relevant responses.

This workflow is practical for knowledge workers, researchers, or support teams who want to quickly extract insights from documents stored in Google Drive, visualize them in a knowledge graph, and interact with the content through AI-driven chat interactions, enhancing understanding and decision-making.