AI Engineer, Product

AI Engineering at Brex is redefining how businesses run their finances by building intelligent, autonomous systems directly into the Brex platform. Our teams develop AI agents that don’t just surface insights—they take action, optimizing spend, managing workflows, and making real-time decisions on behalf of our customers. By deeply integrating proprietary financial data with product and platform infrastructure, we’re turning complex financial operations into simple, automated experiences and setting a new standard for how modern finance works.

You'll be a product engineer building Brex's Audit Agent — an agentic system that reviews customer spend at scale and replaces the manual work traditionally done by BPO teams. The agent itself reasons; the surrounding product harness is what makes that reasoning useful, trustworthy, and operable for real customers.

That product harness is where you'll live. You'll design and ship the workflows, surfaces, and data flows that turn the agent's output into something a reviewer can trust and act on — from the reviewer experience itself, to the contracts with Brex's financial system of record, to the feedback loops that make the agent better over time. You'll move fluidly between backend and frontend as the work demands, picking the right altitude for the problem rather than defending a layer.

This role sits between deep agent design and the customer-facing UI, and it leans toward whichever side the next user problem lives on. You'll partner closely with the engineers driving each — translating ambiguous product requirements into shipped features, pushing back when the design isn't right, and treating the seam between agent reasoning and human workflow as a first-class product surface.

The majority of your work will be backend: system design, data modeling, API shape. A meaningful minority will be on the UI itself, where you'll build alongside our frontend engineers when that's where the next user problem lives.

• →Build and ship customer-facing features in production — from sketch to rollout to iteration based on real customer feedback.
• →Define the data contracts between the agent and the rest of Brex's financial system of record, and evolve them as the product gets sharper.
• →Stand up feedback and evaluation loops that let us quickly gather product signals and close them with product fixes — better signals, better surfaces, better workflows.
• →Run experiments on customer-facing behavior (UI flows, prompt-driven product changes, new review modes) and make calls based on what the data says.
• →Talk to customers and reviewers directly, bring what you learn back into the product, and prioritize what to build next on the team.

This role will be based in our San Francisco office. We are a hybrid environment that combines the energy and connections of being in the office with the benefits and flexibility of working from home. We require three coordinated days in the office per week — Monday, Wednesday, and Thursday. As a perk, we also have up to four weeks per year of fully remote work!

• Strong track record of shipping customer-facing features end-to-end across both backend and frontend — you don't bounce work over a wall, and you've built enough of each to have real opinions.
• Product mindset: you reason about users, workflows, and outcomes first, and treat the system as the means to an end. Comfort with ambiguity and willingness to talk to customers directly.
• Strong bias towards action — you've operated in environments where the next thing to build wasn't handed to you in a ticket, and you've shipped things that didn't exist before.
• Comfort working across team boundaries and pushing back on decisions outside your direct ownership when the seam isn't right — agent design, UX, data model, all fair game.
• Strong backend foundation — system design, data modeling, API shape — and the disposition to ship the UI yourself when that's what the user problem needs, rather than handing it off

• Early-stage startup experience, or time on a small team where you owned a product surface end-to-end.
• Experience building products on top of LLMs or agentic systems — particularly the surrounding harness (evaluation, tracing, feedback loops, human-in-the-loop workflows).
• Background in fintech, compliance, audit, fraud, or other domains where review workflows and traceability matter.
• Experience operating where the underlying system is non-deterministic and the product has to compensate for that.
• Track record of being the engineer teams pull in when a project is stuck across backend, frontend, and a third system that nobody fully owns.