Microsoft · AI-500 · FREE COURSE
AI-500 · Designing and Implementing Multi-Agent AI Solutions
16 video lessons and 146 original practice questions. Follow the course, then put each idea into practice.
YouTube loads only after you press play. Watch on YouTube ↗
Course chapters
16 lessons- 1Decompose goals into agents, workflows, and tools4:03 · Lesson video
- 2Tools, protocols, memory, and model families10:32 · Full-course chapter
- 3Integration, Zero Trust, state, and compute17:10 · Full-course chapter
- 4Observability, monitoring, and developer tools24:26 · Full-course chapter
- 5Advanced prompt engineering and fine-tuning40:19 · Full-course chapter
- 6Context management and agent memory47:28 · Full-course chapter
- 7Multi-agent RAG and knowledge integration54:17 · Full-course chapter
- 8Tool ecosystems and MCP1:01:09 · Full-course chapter
- 9Orchestration, human-in-the-loop, caching, and scaling1:08:25 · Full-course chapter
- 10A2A, frameworks, Transformers, and middleware1:15:39 · Full-course chapter
- 11Evaluation, human review, and continuous improvement1:31:26 · Full-course chapter
- 12Performance, context windows, tokens, and cost1:38:35 · Full-course chapter
- 13Reliability, availability, and tracing in Foundry1:45:13 · Full-course chapter
- 14Security, Key Vault, and AI red teaming2:00:29 · Full-course chapter
- 15Guardrails for multi-agent solutions2:07:35 · Full-course chapter
- 16Release strategies, testing, and CI/CD2:14:23 · Full-course chapter
Practice by domain
Practise the reasoning.
Choose immediate explanations, or complete a mock exam with feedback at the end. The mock questions are also available in everyday practice.
The optional 60-minute timer is a practice setting, not the official exam duration.