The Product team is at the forefront of Brex’s mission to empower employees anywhere to make better financial decisions. With a deep understanding of the business, we identify and scope out the most impactful opportunities for Brex to tackle. We are responsible for aligning cross-functional teams — such as Engineering, Legal, Compliance, and Design — on key decisions. We set strategy and drive products from inception to launch, enabling Brex to grow rapidly and help our customers reach their full potential.
Unlocking AI Agents for the Finance Admin
Companies buy Brex to run their spend, and the person on the hook for that is the finance admin — a controller, an accounting manager, a finance ops lead. Their job comes down to two things they can never quite get ahead of: making sure money is being spent compliantly and documented in time to close the books, and knowing where the company’s money is actually going.
Today they do both by hand, at a scale that doesn’t work. Roughly 4% of travel and entertainment spend is out of policy — for a company spending $10M a year, that’s $400K in violations, duplicates, missing receipts, and overspend. Most companies never find it, because finding it means a human reading transactions one at a time. The ones who do find it either staff a review team or bolt on a separate audit product with its own login. On the other side, the questions that actually matter — where did spend grow this quarter, which team is drifting, what’s driving the number — sit behind a queue for an analyst, or a spreadsheet the admin builds themselves at month-end. Nobody is continuously watching.
The AI team is working to give every finance admin two people they can’t hire today: an auditor and a business analyst.
•
The auditor. An agent that reads a company’s own expense policy and audit rules and reviews every transaction against them — not a 10% sample. It investigates what looks wrong, gathers the missing context by going back to the employee who spent the money, groups what a human should look at, recommends the action, closes what it can safely close, and surfaces the patterns underneath: which policies are constantly broken, which employees are repeat risks, where the real exposure is. Done right, it turns expense compliance from a post-mortem into a first line of defense that runs on every transaction, and shrinks the review pile to the exceptions that genuinely need a person.
•
The business analyst. An agent that answers the questions an admin would otherwise hand to an analyst — asked in plain language against their real spend data — then learns which numbers they care about, turns those into reports their whole company can use, and eventually watches those numbers and tells them when something is off. It starts as a junior analyst who runs queries on request and grows into the analyst for the entire business.
Together they add up to a single answer for the customer: am I in control of my company’s money?
As a PM on Finance Admin AI, you’d take point on one of them with real, end-to-end ownership — which one, and how much of the other you pick up, is something we’ll shape around where you’re strongest and what the team needs when you start.
Both are early and real — in customers’ hands, with the gaps that implies. And no one has a finished playbook for this: what makes an agent’s judgment about a policy trustworthy enough to act on, how you evaluate an answer that has no single correct value, how you keep quality moving up as the models shift underneath you. You’ll invent much of that, with the ownership and pace of a startup and the backing of an established company behind you.