Build AI Agents with smolagents

Course overview

Designing AI agents with smolagents: build tool-enabled agents, implement agentic RAG and multi-step agents, and create multi-agent systems with specialist roles, coordination and agent memory management.

Practical introduction to smolagents that walks through creating agents that perform calculations and tasks without hand-coding every function, and how to attach external tools to agents.

Deep dive into agentic RAG: ordering knowledge sources into answers, what happens inside a RAG pipeline and the step that injects semantic context.

Study of multi-agent systems: designing specialist agents, coordinating an agent team and approaches to manage agent memory and validate inter-agent actions.