AI engineering, opened up

See the mechanism. Run the lab. Know when to stop.

Animated, step-by-step tutorials on context, memory, loops, evals, guardrails and cost, with labs you run on your own machine. No videos: read at your pace and replay each mechanism.

Browse tutorialsFree concepts

Step 1 of every tutorial is free · Labs are plain Python · Nothing to install

eval-engineering/labs/answer

  
  
Real output from the Eval Engineering labs. Precision 1.00, recall 0.59: every answer it flags as wrong really is wrong, but it misses 41% of the wrong answers. Plain Python, runs offline in under a second.










  
Start free

  

    
43 concept explainers
Start with How LLMs Think, then KV cache, reranking, prompt injection… one animated page each.

    
Step 1 of every tutorial
Read the first step of any volume before you decide to buy.

    
Two leadership volumes
For engineering leaders, free to read: The Judgment Gap and Reading an Eval Report.

  





  
Tutorials · 10 volumes

  
Three tracks

  
Foundations
Context EngineeringDeciding what travels with every request, and what gets looked up.Graph EngineeringBuilding a retrieval layer that follows connections instead of matching resemblance.Agent Harness EngineeringThe scaffolding around an agent loop, and what each layer of it buys.Agent ArchitectureWho asks whom, in what order, and what has to be true before that arrangement is worth building.
Reliability
Loop EngineeringHow many attempts, by how many agents, and what has to be true before it stops.Memory EngineeringWhat an agent should write down, what it should re-derive, and what has to be true before a record is trusted again.Self-Learning AgentsWhat an agent may conclude from its own attempts, and what has to be true before it acts on it.
Production
Eval EngineeringYou have a number that says your agent works. This is how to find out whether the number means anything.Agents GuardrailA guardrail’s value is set by where it sits, not by how good it is.Token Cost EngineeringWhich cost reductions cannot cost you an answer, in what order, and when to stop.





  
For engineering leaders

  
Your team ships with agents now. These are the questions you'll be asked.

  
Short, standalone volumes: accountability, measurement, what to automate, where your next senior engineer comes from.

  See the leadership volumes



Latest article
Engineering2026-09-24
What Changes for an SDET: From Writing Tests to Judging Signal
Writing tests was the most automatable thing we do, and it went first.



  
New explainers and articles by email

  
One short email when something new is published.