Internal AI Assistant (RAG)

A complete Internal AI Assistant built end-to-end using n8n workflows and Retrieval-Augmented Generation (RAG) architecture. Uses pgvector for vector similarity search and Ollama for running a local Large Language Model, ensuring data privacy.

Internal AI Assistant (RAG) - 1

My Role

AI Developer — Architecture, Vector DB, LLM Integration, Full Stack

Tech Stack

n8nPythonpgvectorOllamaLLMPostgreSQLRAG

Source Code

github.com/wongsakron

Key Highlights

Full RAG pipeline automated with n8n: document ingestion → embedding → retrieval → generation
pgvector for high-performance vector similarity search
Local LLM deployment with Ollama (no cloud dependency)
Complete data privacy — all processing happens on-premises