Build practical AI skills.
Start with the foundations. Then go deeper into the tools you use.
Three good places to start.
Claude Code
Use Claude Code to plan, build, and ship. Learn how to review its work and keep the project on track.
AI BUILDINGHow LLMs Work
Learn how models, context windows, tools, and agents work.
FOUNDATIONSEvals
Decide what a good result looks like, build an eval, and use it to improve the product.
AI QUALITYStart with the same foundation.
Go deeper by role.
Everyone learns how AI works. Then each role builds the skills it needs.
Decide what to build.
Understand the system, make the tradeoffs, and define what good looks like.
Design how the system behaves.
See what happens behind the interface so you can design better controls, feedback, and recovery.
Build systems that hold up.
Work through requests, retrieval, architecture, and reliability.
Put AI to work.
Use AI tools to improve workflows, handoffs, and day-to-day execution.
Recently published
Build RAG Systems
Design a retrieval-augmented system that prepares evidence, retrieves and ranks it, grounds claims, and exposes failures.
Build AI Prototypes
Turn a written prompt into working software you can click: test ideas, replicate real product UIs, and build tools for yourself.
Scale Software Systems
Systems do not need every scaling tool on day one. By the end you can start with one URL shortener and find its next bottleneck. You can choose the smallest useful change and carry the design through traffic bursts, service boundaries, zones, and global users.