L

Learn AI

Track progress · learn offline

Open

Lesson 1 of 6

The AI Engineer rises

6 min read

Ten years ago, doing anything with AI meant a PhD and a rack of GPUs. Today you make an API call. So who does the work in between?

A job that didn't exist

For decades, AI lived in research labs. Building something meant collecting data, training a model from scratch, and tuning it for months. Then powerful [foundation models](glossary://foundation-model) arrived — general models you can just call — and a gap opened between the researchers who build those models and the developers who ship products. The person who fills that gap is the AI engineer.

The AI engineer sits between research and product: not training models from scratch, but turning ready-made ones into things people use.

Why the role appeared now

It's not that people got smarter — the tools changed. When state-of-the-art AI became an [API](glossary://api) call away, the hard part stopped being training the model and became using it well: prompting it, wiring it to data, evaluating outputs, and shipping something reliable. That's a software job, not a research one — and there are far more software builders ready to do it than researchers.

You don't need to have trained a neural network to be an AI engineer. Most of the job is classic engineering — APIs, data, testing, product sense — aimed at a new kind of component.

Foundation models turned AI from something you train into something you build with. That shift is what created the role.

What the AI engineer does

A team wants to add an AI feature to their app using an existing model. Whose job is that, mostly?

Continue in the app

Take the whole Becoming an AI Engineer course — tracked

Get your personalized path, progress and streaks in the app — this lesson and every next one, in order.