AI Engineer Agentic Track: Agent & MCP Training Program
Build autonomous AI agents and master the Model Context Protocol (MCP) — the complete path from single-agent scripts to production multi-agent systems.
- 28 hours on-demand content
- Full lifetime access
- Access on mobile and desktop
- Certificate of completion
What you'll learn
Description
The AI Engineer Agentic Track is a deep, project-based path into building autonomous AI agents — the kind that plan, use tools, and coordinate with other agents rather than just answering a single prompt. You'll implement the Model Context Protocol (MCP) from scratch to understand exactly how agents connect to tools and data.
By the end, you'll have built and deployed a multi-agent system of your own, with the monitoring and guardrails needed to run it in production.
Curriculum
Agent fundamentals
- What makes a system "agentic" 26 min
- Planning, reasoning, and reflection loops 39 min
- Tool use and function calling 34 min
Model Context Protocol (MCP) deep dive
- MCP architecture and core concepts 41 min
- Building an MCP server 47 min
Multi-agent systems
- Coordinating agents and hand-offs 36 min
- Shared state and communication patterns 33 min
Production agentic systems
- Monitoring and evaluating agents 30 min
- Capstone: shipping your multi-agent system 52 min
Requirements
- Comfortable writing Python; the AI Coder program is a helpful (not required) prerequisite
- A computer with internet access and a code editor
- Basic familiarity with REST APIs is helpful
Instructor
Ed builds and teaches applied AI engineering courses focused on shipping real systems, not just demos — from LLM APIs to full agentic pipelines.