Software Engineer II

We're rebuilding interface.ai as an AI-native company — and the engineering that powers millions of live banking conversations has to be built by people who own what they ship. As a Software Engineer II, you own modules end-to-end: the design, the build, the evals, and the quality in production.

This is not a maintain-someone-else's-code job. The product works, more than 100 financial institutions run on it, and the business is cash-flow positive. The job is to carry real scope inside a deliberately lean engineering org — one that serves millions of conversations precisely because every engineer on it ships like it's theirs. Agents take the routine work; you spend your time on the design decisions and the evaluation harnesses that prove what shipped is correct.

We build in person, from San Francisco. At our stage speed is the whole game, and the fastest path between a problem and a decision is two engineers at the same desk. Scope, title, and pay track what you've put into production — tenure doesn't enter the math.

Your modules. End-to-end ownership — design, build, eval, ship, iterate. You make the design calls within your module and answer for its quality in production.

The ship cadence. Weekly releases to live financial institutions — real customers, real money, real consequences. No demos, no staging purgatory.

The AI leverage. Build with LLMs daily. Steer agents through the routine work so your time goes to the decisions that matter.

The proof. The evaluation harnesses that verify model output. In a regulated industry, "it looks right" isn't a standard — evals are how you know what shipped is correct.

You've designed, built, and shipped production software end-to-end — and owned its quality after launch, not just at launch.

You've built real product with LLMs, and you're rigorous about verifying model output. Vibes don't pass review; evals do.

~2–5 years of professional experience — we level on the scope of what you've shipped, not your title.

High agency — you'd rather be handed a problem than a ticket. You identify what's broken before it's assigned and close loops without reminders.

AI-native by habit — you use frontier AI tools daily to move faster and think sharper, because that's how this company works.

On-site in downtown San Francisco office 5 days a week — we provide commuter benefits if in the general Bay Area, and relocation support if you're not.

Early-engineer experience at a startup — you know what it means to carry scope when the team is small

Open-source contributions or competitive programming

Experience in fintech or banking infrastructure