About this role
App Intelligence Experience starts from the most interesting problems people have with their apps and the information inside them, then works out how intelligence can help. Some of that ships today, connecting apps to Siri; some we prototype for what comes next, in apps and around them. Our work spans iOS, iPadOS, and macOS, and reaches hundreds of millions of people.
As an intern, you will take one of those problems and, with a mentor on the team, build a working prototype of how we could solve it. We prototype in real builds of the OS rather than mockups, so what you make runs on a device and can be put in someone's hands.
You will start from a problem people have with their apps: understanding it, working out what a good answer looks like, and deciding where intelligence can help. Then you will build it end to end, from the model to the interface. That means writing Swift on Apple platforms, working with large language models through prompts, tools, and structured output, and handling the cases where the model gets it wrong. Most of this you will do independently, making your own design calls, with your mentor to turn to. You will measure whether the result helps, demo it to the team and its partners, and write down what worked, what did not, and the questions it raises.
Minimum Qualifications
Programming experience in at least one language such as Swift, Python, C++, or Java, from coursework, research, or
your own projects.
A solid foundation in data structures, algorithms, and object-oriented programming.
At least one project you built from start to finish, such as an app, a tool, or a research prototype.
Preferred Qualifications
Experience building with large language models: prompting, tool or function calling, structured output, and what to do when the model's answer is wrong.
Experience building apps for iOS or macOS in Swift.
Experience measuring whether something you built works, through an evaluation set, a user test, or a benchmark, and changing it because of what you found.
A habit of starting from the problem a person has before reaching for a technology, and an eye for interaction design.
Fluency with AI coding assistants, including a sense of where not to trust them.
Projects from hackathons, game jams, or the Swift Student Challenge.