Stack-wise Concepts
Concepts grouped by the stack you meet them in, rather than alphabetically. Pick a stack to see its concepts; each one opens to a short explanation of what it is and why it matters.
Maths for AI Stack → Linear algebra, calculus, probability and statistics — every concept paired with where it actually shows up in AI, what is safe to skip, and the eight derivations worth being able to do on a whiteboard. 52 concepts Agentic Stack → What an agent is made of — harness, tools, memory, skills — and the runtime, isolation, and evaluation machinery needed to run one in production. 18 concepts API Stack → How services expose and consume each other: protocols and transports, identity and access, and the reliability concerns that decide whether an API survives contact with real traffic. 18 concepts Data Science Stack → The tools you compute with and the statistics that keep the answer honest — from SQL and pandas through hypothesis testing, experiment design, and causal inference. 19 concepts ML Stack → Models and the mechanics that train them: classical algorithms, neural architectures from MLP to transformer, and the losses, optimizers, and regularization underneath. 20 concepts AWS Stack → The twenty-odd services you actually reach for out of hundreds — containers and compute, storage, data and analytics, messaging, and the identity and operations layer underneath all of it. 28 concepts DevOps Stack → Everything between code that works and a service running reliably for other people: pipelines, containers and orchestration, infrastructure as code, release strategies, observability, and the security checks along the way. 31 concepts
Related: Vision & Multimodal AI covers the vision subdomains in the same format, the Toolkit covers the same territory as a flat glossary, and Skills lists the specific tools I work with.