Build agentic AI
across every stack
A hands-on learning platform for engineers, architects, and AI builders. Master the vendor-neutral fundamentals of agentic AI first, then go deep on the ecosystem you ship in — with structured learnings, interactive animations, labs, retrieval practice, and a context-aware assistant on every page.
Master the concepts, vendor-neutral
The fundamentals every agent builder needs, independent of any cloud or SDK — how LLMs work, prompting, tools & MCP, retrieval, memory, agents & multi-agent systems, and the product layer. In depth, with interactive diagrams and self-checks.
8 chapters · self-paced · no prerequisites
Then go deep by ecosystem
Applied training paths that take the fundamentals into a real stack — provisioning, identity, tools, retrieval, evaluation, observability, security, and a production capstone. One path per technological ecosystem, each with hands-on notebooks you run in the cloud.
Live: Microsoft Azure · more ecosystems on the way
Training paths
Start with Core Concepts, then choose the ecosystem you build in. New paths are added over time.
Vendor-neutral foundations of agentic AI.
Build to production on Microsoft Foundry.
Bedrock AgentCore & the Strands SDK
Vertex AI & the Agent Development Kit
Agents SDK · Claude Agent SDK
LangGraph · CrewAI · self-hosted
Understand agents from first principles — tokens, reasoning, tools, retrieval, memory — so the skills transfer to any vendor.
Interactive animations, reveal-to-answer checks, spaced repetition, and notebooks you open in Google Colab.
An account keeps your progress, mastery, and review cards across paths, sessions, and devices.