The posting
You'll build products where receivables teams set direction and AI agents do the follow-through: chasing overdue invoices, resolving payment blockers, and matching incoming cash. You'll take customer problems through working prototypes to production delivery with engineering.
At Causa Prima, we build AI agents that work across company boundaries. The invoice stuck in a buyer's approval queue is also the payment a supplier is chasing. You'll connect those two views so an agent can find out what's blocking payment and help resolve it, instead of sending another reminder nobody can act on.
You may have learned this domain in accounting, audit, advisory, or a receivables team. Here, you'll turn that knowledge into digital products: speaking with customers, building prototypes, and working with developers to make them reliable.
What you'll do
- Find out why invoices stay unpaid. Work with receivables teams to understand how they prioritize overdue balances, track promises to pay, and uncover payment blockers.
- Turn collections knowledge into agent behavior. Design how agents decide which account needs attention, gather the facts, follow up, and escalate. Define what happens when a buyer promises payment, disputes only part of an invoice, or stops replying.
- Resolve disputes across company boundaries. Shape how agents exchange missing documents, investigate deductions, and route discrepancies to the right person.
- Make incoming cash explainable. Design how payments and remittance information are matched to open invoices, including partial payments, missing references, and one transfer covering several invoices.
- Prototype, test, and ship. Build prototypes with AI coding tools, test with users, and work with engineering on production delivery. Turn receivables rules and exceptions into precise specs and acceptance criteria.
- Make the limits and results visible. Define what agents can do within agreed policies, which actions require human approval, and how users trace what happened.
What we're looking for
- Receivables knowledge earned in practice. You've worked on collections, disputes, deductions, or cash application in accounting, audit, advisory, a finance team, or software. Show us the workflows and exceptions you've handled; credentials alone aren't the bar.
- Customer-facing analytical judgment. You've investigated why money remained outstanding beyond the first explanation. You can distinguish an unpaid invoice from a disputed balance or an unapplied receipt, explain the consequences, and change your recommendation when the evidence changes.
- Ownership through delivery. You've taken a financial process or customer problem through to an implemented change, kept people aligned, and resolved the awkward details. You can point to what improved after your recommendation became practice.
- Evidence that you learn by building. You've learned unfamiliar tools to solve real problems and already use AI tools to automate work or build prototypes. Show what you built and how you checked it. Previous product or engineering titles aren't required.
- A bridge from finance to software. You can break a receivables process into steps, data, rules, and exceptions a developer can build against. You learn technical concepts quickly and can explain what an agent should do, when it should stop, and how to check the result.
- Onsite in Munich or Madrid. You're based here or willing to relocate.
- Fluent English. Spanish or German is a plus.
This is a wishlist, not a checklist. If the role excites you but you don't tick every box, we would encourage you to apply anyway.
Nice to have
- Buyer-side process knowledge. You understand how invoice matching, approvals, and payment runs affect when a supplier gets paid, and can use that context to resolve receivables problems.
- ERP implementation experience. You've worked with receivables ledgers, remittance data, or cash-application integrations and dealt with the gap between a documented process and how a customer's system actually behaves.
- Agentic product experience. You've helped turn a customer problem into a digital product. Experience with agents that act within defined policies and human approvals is a further plus.
