For developers, software engineers and tech leads

Become a pro with AI.

Hand-drawn, hype-free lessons on Claude Code, MCP, the Claude API, prompt engineering, evals and local LLMs. Every one comes with code that runs.

Read the blogWatch on YouTube

From the blog

All posts
  1. AI Fluency for developers: the 4D frameworkThe habits behind getting good results from AI, in eight parts.9 parts · Beginner
  2. Building with the Claude API: a hands-on courseFrom your first API call to agents, in fifteen parts.15 parts · Intermediate
  3. MCP in Depth: build MCP servers in TypeScriptBuild one real MCP server, RepoPilot, in nine parts.9 parts · Intermediate
  4. LLM evals and observability: test AI features properlyStop shipping prompts on vibes, in eight parts.8 parts · Intermediate
  5. Local LLMs, hands-on: run models on your own machineRun, measure and use models on your own hardware, in eight parts.8 parts · Intermediate

How every part works

  1. 1

    A concept video

    One idea, drawn live, in plain English. Under ten minutes.

  2. 2

    A hands-on video

    We build it together, from an empty folder, and run it.

  3. 3

    The written lesson

    Here on this site: the explanation and the code to copy.

About DevAIper

What is DevAIper?

DevAIper is a YouTube channel and website that teaches software engineers how to use AI properly: prompt engineering, the Claude API, MCP, evals and local LLMs.

Who is DevAIper for?

Developers, software engineers and tech leads, from junior to staff. If you can read a stack trace, you can follow along.

Is it free?

Yes. The videos, the written lessons and the code are free.

What language are the code examples in?

TypeScript, unless a lesson says otherwise. Every example is run before it is published.

New videos every Tuesday and Friday.

Subscribe and the next part shows up in your feed.

Subscribe on YouTube