AI & Machine LearningIntermediate

Autonomous AI Agents & RAG

Architect production-ready multi-agent workflows, tool-use systems, and vector search retrieval pipelines.

Interactive Live CohortProject-basedComprehensive

What you’ll learn

  • Build production RAG pipelines with chunking, hybrid search, and embeddings
  • Implement multi-agent systems with task decomposition and tool calling
  • Evaluate agent accuracy, hallucination rates, and latency under production conditions

Build as you learn

An autonomous market research multi-agent system. A practical deliverable to help connect the course concepts.

Course curriculum

Module 1: Embeddings & Vector Databases
  • Embeddings, chunking strategies, and similarity metrics
  • Vector databases setup (ChromaDB, Pinecone)
  • Hybrid keyword and dense vector search
Module 2: Advanced Retrieval-Augmented Generation
  • Advanced RAG patterns: query rewriting, reranking, and contextual compression
  • Handling multi-document collections
  • Evaluating retrieval precision and recall with Ragas
Module 3: Function Calling & Tool Execution
  • Function calling conventions and JSON structured outputs
  • Defining external tools and API executors
  • Building error recovery loops for tool failures
Module 4: Multi-Agent Collaboration & Memory
  • Planning, task decomposition, and execution loops (ReAct pattern)
  • State management and agent memory systems
  • Multi-agent orchestration and delegation hierarchies
Module 5: Evaluation, Guardrails & Deployment
  • Production guardrails, sanitization, and output validation
  • Measuring hallucination and latency bottlenecks
  • Securing agent tools against prompt injection
Module 6: Capstone: Autonomous Research Agent
  • Architecting an autonomous research and synthesis agent
  • Integrating web search and database tools
  • End-to-end evaluation, testing, and deployment

Prerequisites

Familiarity with Python or JavaScript and core generative AI concepts.

Keep exploring