Welcome to LLM Agents Academy
SOTA Agents. 14 modules, 101 lessons.
Foundations of LLM Agents
What agents are, why they matter, and the core loop that powers every agentic system
7 lessons
Prompting & Reasoning
Prompt engineering techniques that unlock reliable agent behavior and structured reasoning
6 lessons
Tool Use & Function Calling
Giving agents the ability to act on the world through tools, APIs, and the Model Context Protocol
7 lessons
Agent Architectures
Core architectural patterns from ReAct to workflow agents and how to choose between them
7 lessons
Agent Harnesses and Runtime Systems
The runtime layer that turns agent architectures into executable systems — managing the agent loop, tool dispatch, context and state flow, persistence, observability, error recovery, and human intervention across real production harnesses.
12 lessons
Agent Design Patterns
Reusable software engineering patterns for building robust, maintainable agent systems
7 lessons
Memory & Knowledge
Giving agents short-term context, long-term recall, and access to external knowledge
7 lessons
Agent Frameworks & SDKs
Hands-on tour of the major agent frameworks and when to use each one
7 lessons
Multi-Modal Agents
Building agents that see, hear, and generate across text, images, audio, and video
7 lessons
Multi-Agent Systems
Orchestrating multiple agents that collaborate, delegate, and communicate
7 lessons
Evaluation & Testing
Measuring agent quality, building test suites, and establishing confidence in agent behavior
6 lessons
Production, Deployment & Safety
Taking agents from prototype to production: deployment infrastructure, reliability, safety, cost, and monitoring
7 lessons
Advanced Patterns & Real-World Applications
Cutting-edge agent patterns, specialized domains, and a capstone project
7 lessons
Enterprise Agent Applications
Applying agent patterns to enterprise verticals: data, support, DevOps, workflows, and governance
7 lessons