BastinFlorian/RAG-on-GCP-with-VertexAI

Implements retrieval-augmented generation (RAG) by combining Vertex AI Search's vector embeddings with Confluence data sources, enabling semantic question-answering over proprietary documentation. The architecture extracts and chunks Confluence pages, generates embeddings via Vertex AI, stores them in a vector index, and deploys a Cloud Run chatbot endpoint that retrieves relevant context to ground LLM responses. Infrastructure-as-code deployment via Terraform automates the entire GCP setup including artifact registry, secret management, and index endpoint provisioning.

No commits in the last 6 months.