End-to-End Agentic AI & Automation Lab

The lab progresses from foundational modules on Pydantic data validation and basic LangChain retrieval-augmented generation with FAISS and Pinecone, through LangGraph orchestration covering state graphs, agentic RAG, multi-agent supervisors, human-in-the-loop review and corrective RAG, and AutoGen patterns such as round-robin and swarm collaboration with custom tools.

Later modules cover industry-grade Model Context Protocol implementations for tool execution, web search and Notion integration; production retrieval-augmented generation with hybrid search, BM25, LlamaParse and semantic routing alongside long- and short-term memory; low-code, zero-code multi-agent orchestration with n8n and LangFlow; and fine-tuning with LoRA and Unsloth served through vLLM. It also includes full example products, such as an AI interviewer, an applicant tracking system, and a stateful persistent chatbot, and is licensed MIT.