Learn AI by building with JavaScript

Five hands-on Quests, self-paced and available anytime.

Pause animation

Self-paced. Open source. Available anytime.

Start with Quest 1 Choose a Quest

One path, five practical outcomes

1. Run AI locally Use on-device models for privacy, speed, and offline access.
2. Understand the model lifecycle Explore selection, customization, evaluation, tracing, and AI red teaming.
3. Ground responses with your data Build a RAG pipeline with ingestion, retrieval, citations, streaming, and history.
4. Build and evaluate an agent Compare models, connect MCP tools, test responses, and export working code.
5. Explore a complete agent system Trace how an agent, APIs, MCP services, and application interfaces work together.

Choose your starting point

New to AI development

Begin with Quest 1: Local AI Development, then follow the Quests in order. Each introduces a new pattern while keeping JavaScript or TypeScript at the center.

Ready for retrieval or agents

Go directly to the Quest directory to compare outcomes, environments, and prerequisites before choosing your path.

Learn on your schedule

Every Quest is available whenever you are. Work through the written guide, use the optional recording when a walkthrough helps, and adapt the examples to your own scenario.

Read the start guide

Looking for earlier program results? Visit the program archive.