The Role:We're looking for an AI engineer who wants to help us rewrite the rules of how we build software. You'll work across the GovPilot ecosystem, GovTech, SDL, and Hedgerow, on a team figuring out what comes after the standard agile playbook when AI agents can write, test, and ship code.
This is not a role for someone who wants a defined stack and a backlog. You'll move across languages and codebases as needed. You'll be judged on outcomes, not on hours logged or tickets closed. We want someone who runs fast, sets up their own systems, and pushes AI-assisted development further than we've taken it so far, and does it safely.
You should be someone who is obsessed with AI engineering: you read the papers, try the new tools the week they ship, and have opinions about where this is all going.
What You'll Do:- Build and ship features across the GovPilot, SDL, and Hedgerow codebases, picking up new languages and frameworks as the work demands
- Design and build systems and workflows that let AI agents work at scale without sacrificing safety or quality
- Run your own email, calendar, and task management with a high degree of autonomy. No one is going to hand you a step-by-step plan
- Propose and pilot new ways of working that go beyond standard agile or sprint-based development
- Set up guardrails, evals, and review processes so AI-generated work can be trusted and shipped quickly
- Partner with engineering leadership to figure out what our development lifecycle should look like a year from now, and start building toward it now
What You Bring:- Deep hands-on experience with AI coding tools (Claude Code, Cursor, Copilot, or similar) as a daily driver, not an occasional assist
- Comfort moving across programming languages and stacks without needing ramp-up time
- A track record of working with minimal oversight: you set your own priorities and follow through
- Strong instincts for what can and can't be trusted to run autonomously, and the judgment to build guardrails around it
- Genuine curiosity about how AI changes software development, not just as a productivity boost but as a different way of working
- You must reside in and be authorized to work in the United States
What Makes You Stand Out:- Experience designing or running agentic workflows in production, not just prototypes
- A history of building or proposing new team processes, not just following the ones you were given
- Experience across both legacy and modern systems
- A public track record (blog, open source, talks) of thinking seriously about AI engineering
How to ApplyIf you're the kind of person who's already been experimenting with agentic AI workflows on your own time, and you want to help a team figure this out for real, send us your resume.
