About

The internals, in plain words.

Internals Decoded is a technical publishing studio focused on AI. We take apart the models, agents, and tools shaping the world right now, then explain them deeply enough to be true and simply enough to follow. The whole library is free to read.

The approach

Deep and readable, at once.

Most writing about AI picks a side. It is either rigorous and impenetrable, or friendly and shallow. We try to hold both, so the same piece can meet you whether the topic is new to you or something you already build with every day.

The origin idea

Go under the hood, every time.

The question is always what is really going on inside, from a beginner's first prompt to a production agent system. We show the structures underneath, the tradeoffs behind them, and why things are the way they are, in language you can actually follow.

Who it is for

The curious

You keep hearing these terms and want to understand what is actually happening underneath, not just the headline.

The builders

You work with these tools and want the model behind the model, so the tradeoffs you make are informed ones.

The architects

You design systems and need the reasoning behind a design, not just the recipe, so you can adapt it to your own.

Everything here is free to read, and there is a lot of it. Pick a track and start wherever you are.

Start reading →

Opening the library

The internals, in plain words.

Deep and readable, at once.

Go under the hood, every time.

Who it is for

The curious

The builders

The architects

Keep up with AI. One email a week.

The internals, in plain words.

Deep and readable, at once.

Go under the hood, every time.

Who it is for

The curious

The builders

The architects

Keep up with AI. One email a week.