Learning path
AI Agents and Tool Use
What makes an agent, function calling, LangGraph, CrewAI teams and MCP, with animated films and runnable Python. Part of the Gen AI Developer premium track.
-
1
What Makes an Agent: Plan, Act, Observe
Chatbot, workflow or agent? The five parts of every LLM agent (model, tools, loop, memory, stop rule), a runnable travel-booking agent you can trace step by step, and…
-
2
Function Calling and Tools
Give a real LLM tools: tool definitions with JSON schemas, the tool_use and tool_result messages, a complete Claude tool loop, validating arguments, returning errors, and choosing between parallel…
-
3
Multi-Step Agents with LangGraph
Build agents as graphs: state, nodes, edges and conditional edges, a write-check-revise loop you can run in plain Python, the same agent in LangGraph, the prebuilt ReAct agent,…
-
4
Teams of Agents with CrewAI
Split a job across specialist agents: roles, goals and backstories, tasks with expected outputs, sequential and hierarchical crews, a runnable three-agent crew in plain Python, and what a…
-
5
MCP: Connecting Agents to Apps and Data
The Model Context Protocol in plain words: hosts, clients and servers, tools, resources and prompts, the JSON-RPC messages behind tools/list and tools/call, a tiny server in Python, connecting…