Software Engineering Intern (American University)

• Company: Rigel
• Term: One semester, potential extension
• Open to: American University students

Rigel is an early-stage startup in AU’s Innovation Lab. We’re building Compass, an open-source tool for running a whole team of AI coding agents. The user manages a tree of Manager agents. Each one owns a piece of the product, hands tasks to worker agents, and puts every change through design, testing, and review before a person signs off. We build Compass with Compass, so most of our own code is written by agents and reviewed by us. You can read more at rigel.build.

The Internship

You’ll work directly with the founder, Matt Wilkinson, a 2021 AU grad and experienced software engineer who has worked at Roku, Google, and Capital One. The main goal is for you to learn. Plan on extensive sessions covering software design and architecture, engineering best practices, and how to apply those in the age of AI, especially at scale. We’ll also cover how to review code effectively, especially AI-written code.

The code you write ships. Interns work on Compass and on the infrastructure the company runs on, which covers most of the stack: a Go server, a SolidJS web app, a desktop app, the TypeScript agent runtime, Kubernetes, Pulumi, Nix, Postgres, and CI. You don’t need to know any of it going in.

The work is mostly remote and fits around your classes. Matt will be coming to the DC area occasionally to meet in person. You can also cowork in the Innovation Lab space on campus when needed.

What You’ll Learn

You’ll learn how a professional team ships software. Design docs, small changes that are easy to review, tests, CI, and code review are the habits big engineering orgs depend on, and you’ll pick them up one on one.

You’ll also learn to actually work with AI agents. Prompting a chat assistant is the easy part. Here you’ll plan work for agents, read what they write, catch where they go wrong, and eventually run several at once.

You’ll also see an early startup from the inside, including product decisions and fundraising. What you build can be in the product the same week.

Who We’re Looking For

Your resume matters less to us than how you learn. We’d like to see your GitHub, or something you’ve built, especially a project nobody assigned you. It helps if you’ve picked up a few languages or tools on your own along the way.

You should be curious about AI coding tools, comfortable asking questions, and able to take feedback and come back with a fix.