Architected and deployed a comprehensive Retrieval-Augmented Generation (RAG) pipeline using n8n. This complex workflow automates the ingestion of multiple file formats (PDFs, Excel, CSVs, and Text documents) directly from Microsoft OneDrive. The system automatically extracts, splits, and embeds the document text using OpenAI, storing the vectorized data in a Supabase Vector Store. Furthermore, it integrates a conversational AI agent powered by Anthropic's language model, equipped with persistent memory and document query tools, allowing users to interactively search and retrieve accurate insights from the dynamic knowledge base.

n8n, RAG (Retrieval-Augmented Generation), Workflow Automation, AI Agent, Supabase, Vector Database, OpenAI Embeddings, Anthropic, Document Processing, API Integration.