AI Engineering
Start here: build, test and run an app around an AI model.
View the programGuides
You already build software. An AI model is a new kind of building block: powerful, but it can answer differently each time. This path teaches you to build with it the way you build everything else: tested, measured and released.
What changes
| Compare | Software you know | Software with an AI model |
|---|---|---|
| Testing | Checks that pass or fail | Test sets and scores that move |
| Debugging | Error messages | Reading answers and finding the pattern in the mistakes |
| Building blocks | Libraries that always behave the same | Models that change with each version |
| Running cost | Mostly fixed | You pay for every request to the model |
| What stays | Version control, APIs, deployment, code review | All of it |
Yes, and they start ahead. AI engineering is mostly software engineering with an AI model in the middle. You already know version control, testing, APIs and deployment. You add three things: how models behave, how to test answers that change, and how to keep every request fast. The quickest way to learn is to build one app from start to finish.
Where to start
Start here: build, test and run an app around an AI model.
View the programTake this first only if you want to learn how models work inside.
View the programLater, to make those apps work at customers.
View the programWhat it takes
Questions