RAG & Vector Database Lab

Overview

The RAG & Vector Database Lab is your sandbox for building knowledge-grounded AI applications. When an LLM needs to answer questions about your company's documents, customer data, or proprietary knowledge — that's RAG, and this is where you learn to build it. The lab provides pre-deployed vector databases (ChromaDB and Weaviate running on GKE), access to embedding APIs (Vertex AI and open-source models), document processing pipelines, and evaluation frameworks. You'll ingest real documents, build search pipelines, evaluate retrieval quality, and wire everything together into production-ready applications.

What You'll Do in This Lab

• Ingest thousands of documents — PDFs, HTML, Markdown — with automated parsing and chunking
• Generate embeddings using Vertex AI Embeddings API and open-source models
• Store and query vectors in ChromaDB and Weaviate with metadata filtering
• Build multi-step RAG pipelines with query decomposition and re-ranking
• Build AI agents with tool use, memory, and multi-step reasoning
• Evaluate RAG quality with RAGAS — faithfulness, relevance, and answer correctness

Lab Workflow