Summary

• Causa Prima has raised $10m in pre-seed funding led by Creandum.
• The company wants buyers’ and suppliers’ AI agents to resolve invoice disputes, negotiate payment terms, and capture early-payment discounts.
• The startup is trying to build a shared network between companies, rather than automating only one side of accounts payable or receivable.

Causa Prima has raised $10m in pre-seed funding to build a B2B payments network where AI agents act for buyers and suppliers, handling invoice disputes, payment terms, approvals, and early-payment negotiations.

The round was led by Creandum, with participation from Kfund, HelloWorld, Angel Invest, and angel investors from companies including Qonto, Pennylane, SAP, ING, SoFi, Lidl, and DeepMind. The company is based in Madrid and Munich, and was founded by Maex Ament, Henrik Gebbing, and Philip Stanislaus.

Causa Prima is targeting a stubborn part of enterprise finance. Accounts payable and accounts receivable software has automated parts of invoice processing, matching, approval, scheduling, and reconciliation, yet much of the friction sits between companies rather than inside one finance stack.

When an invoice is disputed, a purchase order does not match, a payment term is renegotiated, or a supplier wants early payment, the process often falls back into email, spreadsheets, calls, and manual exception handling. Finance software can accelerate individual steps, but the handoff between buyer and supplier still absorbs time.

The startup’s answer is a shared network where each side can run AI agents that negotiate and resolve issues. In the company’s model, agents representing a buyer and a supplier can compare evidence, propose settlement terms, agree an early-payment discount, or clear a disputed invoice without waiting for human teams to trade messages for weeks.

That proposition is ambitious because B2B payments are not merely workflow. Payment timing is tied to working capital, treasury strategy, supplier relationships, fraud controls, procurement rules, ERP systems, and legal obligations. A system that changes payment outcomes must be trusted by both sides of the transaction.

The founding team gives the company relevant market experience. Ament co-founded Taulia, the working-capital and supply-chain finance platform acquired by SAP. Gebbing co-founded Finoa, the German digital-asset custody company. Stanislaus has an enterprise security background from Oak Security. The category requires all three disciplines: payments knowledge, network building, and security control.

Finance teams are under pressure to reduce manual back-office work, improve cash visibility, and strengthen controls while payment fraud, supplier stress, and macroeconomic uncertainty remain live concerns. Vendors across procurement, ERP, treasury, and payment software are adding AI features, but many of those tools still sit inside one company’s systems.

Causa Prima is betting that agents can operate across organisational boundaries. That changes the commercial challenge. Network businesses need enough buyers and suppliers to make participation useful, and large companies will move slowly before giving software agents authority over anything that affects cash, supplier relationships, or contractual terms.

Governance will decide whether the model can scale. Finance leaders may accept an agent that drafts responses or recommends a discount. Allowing agents to settle disputes, alter payment terms, or trigger financial outcomes requires policy boundaries, approval thresholds, audit logs, human override, and integration with existing controls. Regulated sectors will also need the decision process to stand up to internal audit and external scrutiny.

The potential productivity gain sits in repetitive exceptions. Late payments, invoice queries, missing documents, duplicate records, tax-field mismatches, delivery discrepancies, and early-payment discount requests consume finance time without adding strategic value. If agents can handle bounded cases while escalating genuinely complex exceptions, the reduction in working-capital friction could be tangible.

Europe is a useful market for the thesis because business payments remain fragmented across countries, banks, ERP systems, tax rules, and payment practices. The EU’s push toward instant payments and digital finance may also make companies more willing to revisit the assumptions behind net-30, net-60, and net-90 cycles. Those terms partly reflect human processing speed. Agentic coordination could challenge that logic, provided the trust layer is strong enough.

Causa Prima still has to prove adoption, governance, and network density. The funding round shows where enterprise AI is moving next: away from internal productivity tools and into inter-company workflows where automation affects cash, negotiation, and market infrastructure.