Understand it. Build it.
Make it work.
Practical explanations, small projects and debugging guides for your next step in AI. Start without code, then go deeper as you build.
10 articles
AI, ML and LLMs: a practical starting point
Understand the vocabulary, choose a small first exercise, and learn what you do—and do not—need to build an AI application.
Read the guide
Write prompts you can test
Turn a vague request into a small, repeatable experiment with context, output requirements and a checklist. No coding required.
Read the guide
Your first AI app: one task, one model
Build a small note-to-checklist application before adding agents, tools or a complicated framework.
Read the guide
RAG: from documents to checkable answers
Learn retrieval-augmented generation by building a tiny document question-answering exercise and inspecting its evidence.
Read the guide
Build your first AI evaluation set
Use a small repeatable test set to compare prompt changes and catch regressions before you trust a demo.
Read the guide
Auto-compact is eating your session
Claude Code summarises your conversation when the context fills up. The summary is lossy by design, which is why the model forgets a constraint you set an hour ago. Compact at task boundaries instead.
Read the guide
Why subagents produce shocking bills
Reports of five-figure Claude Code sessions have one shape in common: parallel subagents, each with its own context, and nothing that stops them. Here is the mechanism and the ceiling to put on it.
Read the guide
Your prompt cache expires faster than you think
Cache lifetimes are measured in minutes, not hours. If you think between prompts, you are paying full price for the same prefix over and over — and the fix is about rhythm, not configuration.
Read the guide
A GitHub issue is untrusted input
Two 2026 advisories describe the same shape: an agent reads attacker-controlled text and treats it as instructions. If your agent can read the internet and reach a secret, that is the whole vulnerability.
Read the guide
Reading the docs is not studying
The common report from Claude certification candidates is that they prepared by reading and were tested on judgement. Retake waits make that an expensive way to find out. Here is a cheaper order to do it in.
Read the guide