AI Agents & Automation

Agentic AI is where models stop answering and start acting. Courses here cover agent architectures and reasoning loops, tool and function calling, memory and state across long tasks, multi-agent coordination, and the protocols and frameworks that have emerged around them including MCP, LangGraph, CrewAI and AutoGen. Considerable attention goes to what makes agents fail in production: error recovery, cost runaway, unbounded loops and the guardrails needed before an agent is allowed to take real actions. This is the newest subcategory on the site and the one where course quality varies most, so check publication dates.