Your missionWe already move the money. The harder problem sits on either side of the payment: the finance work that decides what to bill, what to chase and what to reconcile. For most of the businesses we serve that work is manual, spread across systems that do not talk to each other, and visible to finance only after the fact. We are building the AI layer for it.
This role has a forward deployment component, and we mean it literally. You will sit inside our clients' finance teams, learn how the work actually happens, and design workflows that fit their process. Engineers build the connections into their systems. You decide what moves between them, which parts of the work an agent takes, and where a person stays in the loop.
We are early in this product line. The first clients are how we find out what can be generalized and productized, and you will be the person making that call. If that is the interesting part of the job to you, this is your role. If you are looking for a mature roadmap and a settled platform, it is not.
Hybrid, 3 days in office.
Outcomes - your first 12-18 months- Get an agent into daily use inside a client's finance team within your first two quarters. One workflow, doing work a person did by hand the week before.
- Make the data flow a design rather than a diagram. For each workflow, know where the truth lives, what each system owns, what goes stale and how fast, and where today's handoffs drop information on the floor. Engineering builds the connections; you decide what has to move, when, and why.
- Draw the line between AI and human, workflow by workflow. Where the agent acts on its own, where it proposes and a person confirms, where it must stop and escalate. Write it down, defend it to a controller, and move the line as the system earns trust.
- Standardize a process without breaking it. Five people doing the same job five ways becomes one way the product supports, plus the exceptions that genuinely deserve to stay exceptions.
- Shorten the delay between something happening upstream and finance knowing about it. Pick the measurable version of that, baseline it honestly, and move it.
- Turn client-specific work into product. For every workflow, make the call: bespoke, configurable, or core, and be able to defend it.
Also on your plate- Requirements gathering as a craft. Running sessions with controllers, AR and AP leads, billing and collections teams, and leaving with a process map they correct rather than politely accept.
- Writing the specs and flows engineers build from, and staying close enough to the build to answer questions in hours rather than sprints.
- Putting a working prototype in front of a client instead of a slide. You will build those yourself with AI tools.
- Being the person the client's finance lead calls. You own that relationship rather than handing it to an account manager.
- Working through the exceptions instead of around them. Disputes, rate differences, short payments, charges that surface months late, credits and partial settlements are the workflow, not the edge case.
- Partnering with our in-house logistics expertise, and helping shape what we charge for and how.
What you bring- 5+ years building or delivering software for finance and back-office teams, including at least one 0 to 1 you took from nothing to real usage. Product management is the most common path into this role, but we will read implementation leads, solutions leads and finance operators who moved into product with equal interest. We care more about trajectory than title.
- You understand the finance function from the inside: AR and AP, billing, collections, credit control, payment application, reconciliation and the close. You know why a controller cares about an aging bucket and what makes a month end go badly.
- You can read a process and see the data underneath it. Given two systems and a handoff between them, you can work out what each one owns, where the truth lives, what goes stale, and what breaks when a person is the integration.
- Judgment about where AI belongs and where it does not. You can look at a finance workflow and say which parts an agent should own, which parts need a person, and which parts should stay deterministic because control matters more than cleverness.
- Hands-on with AI, in the way a product person is. You have built with LLMs yourself, even if only in prototypes, you are comfortable with prompting, retrieval and evaluation, and you can tell a demo that works from a system that works. You do not need to write production code.
- You are good in a room with a customer. You can run a working session with people who have been burned by software before, and leave with them on your side.
- You can hold your own with engineers without doing their job. You ask good questions about what a system can and cannot give you, and you know when "that's hard" means hard and when it means tedious.
- Working Spanish is a plus. Our partner finance teams work in Spanish, and the signal you need is often in what they say to each other rather than in what gets translated for you. Not a gate if everything else is there.
- Based in Miami, NYC, or Palo Alto, and willing to travel on site with clients in the US and Latin America.
How you work- You go and look. A day sitting with a collector beats a well-run interview, and you would rather do both.
- High agency, high urgency. You decide and move rather than waiting for a room to converge.
- You say no to some of the wish list items. Design partners will give you twenty ideas. You ship the one that moves the number, and you can explain that choice to their CFO.
- You write. Specs, decisions, and what you learned this week, so the rest of the company can build on it.
- You hold a date. When something is going to slip you say so early and propose a smaller version rather than a later one.
- You would rather ship a narrow thing that works than a broad thing that demos.
Nice to have- You were an accountant, controller, auditor or analyst earlier in your career.
- Order-to-cash, AP automation, spend management, billing or revenue systems experience.
- You have shipped a product with a significant LLM component.
- You have built for a complex, multi-party industry where documents and exceptions drive the work.
- Consulting or solutions delivery experience, particularly if you have since moved to the product side and know where the two disagree.
- Experience in Brazil or Mexico, or building for Latin American SMBs.
Even if you don't meet 100% of the qualifications, we encourage you to apply.
Compensation & benefitsThe expected base salary range for this role is $180,000 to $220,000 per year in New York or Palo Alto, and $165,000 to $195,000 per year in Miami. Actual compensation will depend on experience, skills, and qualifications.
- Equity
- Medical, dental and vision
- 15 days paid time off
- Learning and development budget
Hybrid, 3 days in office.
