Staff Software Engineer

CI/CD has been solved. Now it's being reinvented. AI agents don't just run pipelines. They write code, trigger builds, interpret failures, and propose fixes. The interfaces that developers use to understand and trust that work don't exist yet. We're building them.

At CircleCI, our guiding principles aren't posters on a wall. We learn fast: we run experiments, accept failure, and scale what works. We think like the customer: every decision connects back to the developer sitting in front of a failing build at 2 am. We own the outcome: not just our slice of it, but the whole thing. If that's how you already work, you'll fit here.

We don't hire for a team. We hire for CircleCI. Where you land depends on where your strengths create the most leverage, and we figure that out together. You might own the CI pipeline experience, deploy and release tooling, notification infrastructure, monetization flows, AI agent surfaces, or the foundational web platform. The through line across all of it is the same: full-stack ownership, a high bar for craft, and work that directly shapes how tens of thousands of developers experience CircleCI every day.

• Build AI-natively. AI is how we work. It's not a productivity layer on top of engineering, but the default starting point for how problems get approached and solved.
• Own technical strategy for your domain. You decide when and how the architecture evolves, balancing technical debt against velocity. When a project spans teams, you're the one who breaks down the ambiguity and takes ownership of whether it lands.
• Invent new interfaces. Software delivery is being reinvented. Pipelines run autonomously, generate fixes, and surface results that need human judgment. The tools developers use to understand, trust, and act on that work don't exist yet. You'll define what they look like and establish the patterns the team builds on.
• Own the full stack, at scale. Build features end-to-end in Go, React, and TypeScript. Your ownership extends past your own code to the systems and teams around it.
• Connect engineering to the business. Partner directly with Product and Engineering Management to translate business goals into technical strategy. You understand the "why" behind the roadmap and use it to make tradeoffs.
• Champion observability. Drive monitoring and alerting strategy across your domain, not just your own services. Use Honeycomb, Datadog, and Rollbar to reduce noise and improve the org's ability to diagnose problems quickly.
• Mentor at scale. Mentor multiple engineers at once. Build the systems and documentation that transfer knowledge and remove single points of failure, not just your own.

• AI-native engineering practices. You use AI tools daily and can speak to what you've built, where they changed your approach, and where they fall short.
• A track record of shipping at scale. You've built and owned systems that other engineers build on. 8+ years of production full-stack experience, with a history of taking accountability for outcomes.
• Full-stack depth. Strong across the stack — React and TypeScript on the frontend, Go on the backend, owning API design, data modeling, and service integration.
• System design, across domains. You own architecture decisions that span multiple systems and teams. You make the tradeoff calls, document the reasoning, and are the person others escalate to when the answer isn't obvious.
• Business acumen. You understand how your domain connects to the broader business, and you use that context to guide technical decisions, not just implement them.
• Technical leadership. You mentor engineers, give direct feedback on technical direction, and raise the quality bar across teams through how you engage in reviews and design discussions, not just what you ship.

You've stopped waiting for AI to mature and are already building with it, not as a productivity trick, but as a core part of how you think about problems. You care about the product as much as the code. You're drawn to work where the patterns don't exist yet and the best answer requires you to invent it. If that's how you already operate, we want to talk.