Chapters
01
AI Evals
You can't tell if it works by looking at it. Evals are how you know —
six concepts, six live components, from graders to pass@k.
02
Feedback Loops
Your agent retried and failed again. Without a feedback loop, retrying is just hoping — here's what iteration actually means.
03
Agentic System Design
Your multi-agent system ran to completion and returned a confident wrong answer. Silent failures are the default — here's how to design systems that fail visibly.
04
Agent Harness Engineering
It looked done in 20 minutes. It didn't work. The harness around the model — not a smarter one — is what closes that gap.
05
RAG
Hallucinations are a retrieval problem. Grounding your AI in real data is how you fix it.
06
Fine-tuning
Generic model, generic output. Fine-tuning is how you make AI care about your specific problem.
07
MCP Sampling
You gave your server its own API key so it could call the AI directly. The protocol's own fix for that got deprecated within a year — here's why.