How We Build at Roe
We're a small team building AI agents that investigate financial crime for risk and compliance teams. Our customers include companies like eBay, Affirm, and Crossmint. The work is technically hard, the stakes are real, and we're hiring engineers who want to build at the frontier of what's possible with AI.
Before you apply, you should know how we think about engineering. Not our tech stack or our interview process. How we actually work.
Product-Driven Engineering
Every engineer at Roe is a product thinker. We don't have a product team that hands specs to engineering. Engineers understand the why behind what we build and actively shape what we're building next.
This means you need to care about the problem, not just the code. You'll talk to customers. You'll watch them use what you built. You'll have opinions about what we should build next, and those opinions will matter.
If you want to be handed tickets and left alone, this isn't the right place.
Customer-First Mindset
Every engineer at Roe is customer-facing. You'll join customer calls. You'll read their Slack messages. You'll hear their frustrations directly, not filtered through three layers of account management.
This isn't optional. The best product decisions come from engineers who deeply understand the problem space. The only way to deeply understand the problem space is to talk to the people who live in it.
Our customers are compliance teams at financial institutions. They're smart, they're under pressure, and they know their domain better than we ever will. Your job is to listen, translate what you hear into product improvements, and ship those improvements fast.
Extreme Ownership
Engineers at Roe own their areas end-to-end. Design, implementation, testing, deployment, on-call, long-term reliability. There's no separate team to throw things over the fence to.
This is harder than working in a larger organization where responsibilities are sliced thin. It's also more satisfying. When something works, you built it. When something breaks, you fix it. When customers love a feature, you were in the room when they told us.
Ownership means you care about outcomes, not just outputs. You're not done when the PR merges. You're done when the thing works in production and customers are getting value from it.
Move Fast, Validate Faster
If you have a good idea, prototype it. Today. Demo it to the team tomorrow. No approval required, no design review, no committee.
We're an eight-person company competing in a market where speed is everything. The best ideas often come from engineers who see something in the data or notice a pattern in customer feedback. We want those ideas built and validated before anyone has time to talk themselves out of them.
This requires judgment. Not every idea is good. But the cost of trying a bad idea quickly is much lower than the cost of slowly building the wrong thing.
Leverage AI Aggressively
We build AI products, and we use AI to build them. We invest in tools like Cursor, Vercel, and Claude Code, and we expect engineers to use them extensively. If you're not using AI to accelerate your work, you're leaving leverage on the table.
This is a pivotal moment. The engineers who figure out how to work effectively with these tools will build 50x what they could have built two years ago. We want people who are excited about that, not skeptical of it.
Reviews Over Raw Code
Here's the counterintuitive part: when AI can generate code quickly, code review becomes more important, not less.
The bottleneck shifts from writing code to ensuring the code is correct, maintainable, and aligned with how the system should evolve. Thoughtful review catches the subtle bugs and architectural drift that fast iteration can introduce. We put more weight on your ability to review code critically than your ability to produce it quickly.
A Day at Roe
It's Tuesday, which means it's an in-office day. You get in and do a few hours of heads-down work before lunch arrives, then we eat together as a team with no laptops or phones out. We talk about work, mostly: a thorny bug someone's stuck on, a new request from a customer, a pattern someone noticed in the data. There's usually enough going on that we don't run out of things to discuss, though we talk about our lives outside of work too.
In the afternoon, you might meet with a customer to show them what you've been building and get their feedback directly. Maybe some code review, a 1:1 with our head of engineering, or a working session with a teammate on a tricky problem.
You go home in the evening and keep working, not because you have to, but because you want to. The problems are interesting, the progress is visible, and the work matters.
Every week, we show demos and share sales progress openly. Everyone knows where we are, what's working, and what's next.
Why Now
AI agents are becoming capable of real work. Not demos, not prototypes, but production systems that handle complex, high-stakes tasks autonomously. We're building some of the most sophisticated agents in production today, investigating financial crimes that would take human analysts hours to review.
The next few years will determine which companies figure out how to build reliable, trustworthy AI systems and which ones don't. We think we're ahead, but staying ahead requires hiring engineers who are energized by this moment and want to push the boundaries of what's possible.
Open Roles
We're hiring in San Mateo:
Founding Software Engineer, AI Agent (Hybrid)
Founding Staff Engineer, AI Agent (Hybrid)
Solutions Engineer (Hybrid)
If this sounds like the way you want to work, we'd love to talk. Send us a message and your resume to hiring@roe-ai.com!
