Skip to content
W D
GoBD-compliant §203 StGB-compliant Q1

Cash Application Agent

Read incoming payments, assign to debtors, clear invoices - automatically reconciled.

Reads bank statements (CAMT.053, MT940), assigns payments to debtors and open invoices, validates early payment discount deductions and creates clearing entries. Unassignable payments go to the clerk.

Score Dashboard

Agent Readiness 79-86%
Governance Complexity 18-25%
Economic Impact 76-83%
Lighthouse Effect 21-28%
Implementation Complexity 28-35%
Transaction Volume Daily

What This Agent Does

Cash application is the process that assigns incoming payments to the correct customers and open invoices. For clear-cut payments (exact amount, reference number) this is trivial. For deviating payers, partial payments, overpayments or missing references it becomes complex.

The Decision Layer breaks payment assignment into seven decision steps. The bank statement is parsed rule-based. Assignment to debtor and invoice is two-stage - exact match on reference and amount, AI-assisted when deviating. Discount deductions are validated against contract data. Unassignable payments and difference clarifications go to the clerk.

The result: standard payments are automatically assigned and cleared. Reconciliation happens daily instead of weekly. Open items are updated in real time. And the clerk only handles the genuine special cases that require human interpretation.

Micro-Decision Table

Human
Rules Engine
AI Agent
Each row is a decision. Expand to see the decision record and whether it can be challenged.
Read bank statement Are all transactions correctly parsed? Rules Engine

Parsing of CAMT.053 or MT940 format

Decision Record

Rule ID and version number
Input data that triggered the rule
Calculation result and applied formula

Challengeable: Yes - rule application verifiable. Objection possible for incorrect data or wrong rule version.

Assign debtor (exact) Which debtor does this payment belong to? Rules Engine

Exact match on account number or reference

Decision Record

Rule ID and version number
Input data that triggered the rule
Calculation result and applied formula

Challengeable: Yes - rule application verifiable. Objection possible for incorrect data or wrong rule version.

Assign debtor (deviating) Who paid when the payer differs from the debtor? AI Agent

AI assignment for deviating payer or missing reference

Decision Record

Model version and confidence score
Input data and classification result
Decision rationale (explainability)
Audit trail with full traceability

Challengeable: Yes - fully documented, reviewable by humans, objection via formal process.

Assign invoice Which open invoice is being cleared? Rules Engine

Amount plus reference yields unique assignment; partial payment by AI

Decision Record

Rule ID and version number
Input data that triggered the rule
Calculation result and applied formula

Challengeable: Yes - rule application verifiable. Objection possible for incorrect data or wrong rule version.

Validate discount deduction Is the early payment discount deduction justified? Rules Engine Vendor

Check against contract data and discount deadline

Decision Record

Rule ID and version number
Input data that triggered the rule
Calculation result and applied formula

Challengeable: Yes - rule application verifiable. Objection possible for incorrect data or wrong rule version.

Challengeable by: Vendor

Difference clarification How to handle under-, over- or duplicate payment? Human

Human interpretation for payment differences

Decision Record

Decider ID and role
Decision rationale
Timestamp and context

Challengeable: Yes - via manager, works council, or formal objection process.

Unassignable payments What happens with payments without assignment? Human

Human research and clarification required

Decision Record

Decider ID and role
Decision rationale
Timestamp and context

Challengeable: Yes - via manager, works council, or formal objection process.

Decision Record and Right to Challenge

Every decision this agent makes or prepares is documented in a complete decision record. Affected parties (employees, suppliers, auditors) can review, understand, and challenge every individual decision.

Which rule in which version was applied?
What data was the decision based on?
Who (human, rules engine, or AI) decided - and why?
How can the affected person file an objection?
How the Decision Layer enforces this architecturally →

Prerequisites

  • Bank connection for electronic bank statements (CAMT.053 or MT940)
  • ERP system with accounts receivable and open items
  • Customer master data with payment terms and discount agreements
  • Defined tolerances for payment differences

Governance Notes

GoBD-compliant §203 StGB-compliant

GoBD relevance: high - payment assignment is the basis for correct open items and thus for the balance sheet. Incorrect assignment leads to incorrect receivable balances. Discount validation is tax-relevant (reduction of the assessment base). The two human decisions (difference clarification, unassignable payments) are correctly placed: here, human interpretation is genuinely required.

§203 StGB-relevant data is encrypted end-to-end and never passed to AI models in plain text.

Process Documentation Contribution

The Cash Application Agent documents: the bank statement read in, the debtor assignment (exact or AI-assisted), the invoice assignment, the discount validation and the clearing entry. Unassignable payments are documented with the clarification history.

Infrastructure Contribution

The Cash Application Agent builds the incoming payment infrastructure. The bank statement parsing logic (CAMT.053, MT940) is reused by treasury agents. The discount validation complements the Payment Run Agent on the AP side. The difference clarification pattern becomes the standard for all agents handling reconciliation differences.

Does this agent fit your process?

We analyse your specific finance process and show how this agent fits into your system landscape. 30 minutes, no preparation needed.

Analyse your process

Frequently Asked Questions

Which bank formats are supported?

CAMT.053 (ISO 20022) and MT940 (SWIFT). CAMT.053 is the European standard and is preferred. MT940 is supported for legacy systems. The parsing logic is format-independent - new formats can be added.

What is the automatic assignment rate?

Typically 75-90% for B2B payments with reference numbers. The rate depends on payer discipline. For automatic payments (direct debit), the rate is close to 100%. The remainder is presented to the clerk with AI suggestions.

What happens with duplicate payments?

The agent recognises duplicate payments automatically and routes them for difference clarification. The clerk decides whether to initiate a refund or offset the overpayment. The decision is documented.

What Happens Next?

1

30 minutes

Initial call

We analyse your process and identify the optimal starting point.

2

1 week

Discover

Mapping your decision logic. Rule sets documented, Decision Layer designed.

3

3-4 weeks

Build

Production agent in your infrastructure. Governance, audit trail, cert-ready from day 1.

4

12-18 months

Self-sufficient

Full access to source code, prompts and rule versions. No vendor lock-in.

Implement This Agent?

We assess your finance process landscape and show how this agent fits your infrastructure.