RAG, AI Agents & MCP

An advanced AI course: RAG with embeddings and vector databases, AI agents, the Model Context Protocol (MCP), evaluation and safe production use.

Advanced · 13 lessons

Retrieval-Augmented Generation (RAG)

  1. What Is RAG (Retrieval-Augmented Generation)? - Why models need your data, how RAG works end to end, and when to use it instead of fine-tuning.
  2. Chunking, Embeddings and Vector Search - Split documents well, create embeddings, and retrieve by meaning using cosine similarity.
  3. Vector Databases: pgvector, Pinecone, Qdrant, Chroma - What a vector database does, the main options, and how to choose one for your stack.
  4. Advanced RAG: Hybrid Search, Reranking and Citations - Improve retrieval with keyword plus semantic search, reranking, query rewriting and source attribution.

AI agents

  1. AI Agents Explained - The difference between workflows and agents, the agent loop, memory, planning and typical failure modes.
  2. Multi-Agent Systems, Frameworks and the Claude Agent SDK - Supervisor and worker patterns, handoffs, checkpoints, tracing, and how to choose between frameworks and custom code.

Model Context Protocol (MCP)

  1. What Is MCP (Model Context Protocol)? - What MCP is, how hosts, clients and servers fit together, and what tools, resources and prompts mean.
  2. Build an MCP Server in Node.js - A small MCP server with one tool, how to test it with the MCP Inspector and how to connect it to a client.
  3. MCP Security: Permissions, Consent and Prompt Injection - How to connect AI to tools safely: least privilege, OAuth, consent and treating tool output as untrusted.

Evaluation, safety and production

  1. Evaluating LLM Applications - Build evaluation datasets, measure correctness and faithfulness, use LLM-as-a-judge carefully and catch regressions.
  2. AI Safety, Privacy and Responsible Use - Hallucination mitigation, PII protection, bias, moderation, human review and how Anthropic approaches safety.
  3. Running AI in Production: Cost, Reliability and Monitoring - Architecture and operations for AI features: provider abstraction, caching, quotas, tracing, fallbacks and incident response.
  4. Your AI Learning Roadmap and Project Ideas - A suggested order for learning generative AI, and five projects, including a resume and ATS tool, to build a portfolio.