21 sprints organized into 6 tracks. Each one takes about 15 minutes. Pick any track to start.
Start here - master the essential building blocks
Break complex tasks into sequential steps where each output feeds the next
Teach agents to make decisions and direct queries to the right handlers
Run multiple tasks simultaneously for faster, more efficient agents
Enable agents to critique and refine their own outputs
Level up with tools, planning, and collaboration
Connect agents to APIs, databases, and external services
Formulate multi-step strategies through autonomous task decomposition
Coordinate multiple specialized agents to solve complex problems
Give your agents context and retrieval powers
Maintain context across interactions for coherent conversations
Enable agents to improve through feedback and experience
Standardize agent access to data sources through MCP
Augment agents with external knowledge using RAG techniques
Build robust, production-ready systems
Build resilient systems that handle errors elegantly
Integrate human oversight for critical decisions
Implement guardrails for responsible AI behavior
Deploy, monitor, and optimize at scale
Define objectives and monitor progress toward complex goals
Enable seamless communication between different agent systems
Optimize performance while managing costs
Implement chain-of-thought and tree-of-thought strategies
Monitor agent performance in production environments
Manage queues and allocate resources optimally
Enable agents to find novel approaches autonomously
Quick references and deep dives
Advanced techniques for maximum LLM performance
How agents interact with interfaces and the physical world
Quick guide to major agentic frameworks
Build agents with Google's AgentSpace platform
Create command-line agents for developer workflows
Explore advanced reasoning architectures
Agents that write, debug, and refactor code
