Update date not checked yet
Agent Engineering: Building Multi-Agent Systems
A production-focused agent course that builds a research agent and a writing workflow with MCP, then covers evaluation, observability, CI/CD and deployment.
Our take
Goes well past toy agents into evals, observability and deployment, taught by Paul Iusztin; it assumes real engineering experience and costs more than most self-paced agent courses.
Best for: Developers who already call LLM APIs and want to build, test and deploy multi-agent systems.
Based on the course page, syllabus and pricing, not on completing the course. How we rank
Tools and models covered
- Python
- MCP
- Docker
- Opik
Based on the course content as of an unknown date.
Good to know
50% student discount on request. 30-day full refund.
What you'll learn
- Context engineering
- Tool use and planning
- Multi-agent orchestration
- MCP servers
- Agent evaluation
- Observability and deployment
Before you start
- Python (functions, classes, API calls), experience with the OpenAI or Claude APIs, basic Docker and cloud deployment.