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

Three-Way Matching Agent

Purchase order, delivery note, invoice - automatically matched, discrepancies routed.

Compares incoming invoices with purchase orders and goods receipts. Checks quantities, prices and tolerances. Routes approvals by amount and discrepancy - fully rule-based.

Score Dashboard

Agent Readiness 89-96%
Governance Complexity 16-23%
Economic Impact 81-88%
Lighthouse Effect 21-28%
Implementation Complexity 21-28%
Transaction Volume Daily

What This Agent Does

Three-way matching is the reconciliation of purchase order, goods receipt and invoice. Do quantities, prices and terms match? In manual processing this is time-consuming and error-prone - especially with high order volumes and many line items.

The Decision Layer breaks the matching process into six decision steps - all fully rule-based. Purchase order and delivery note are identified via database matching. Quantities and prices are compared numerically, tolerances applied, discrepancies evaluated and routed per approval matrix.

The result: the Three-Way Matching Agent is the ideal starting candidate - zero AI decisions, zero human decisions in the standard case. All six decision steps are deterministic. The error rate in invoice verification drops drastically, and throughput time from days to minutes.

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.
Find purchase order for invoice Which purchase order belongs to this invoice? Rules Engine Vendor

Database match on PO number and vendor

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

Find goods receipt Which delivery note or goods receipt belongs to it? Rules Engine

Database match on PO 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.

Quantity matching Do ordered, delivered and invoiced quantities match? Rules Engine Vendor

Numerical comparison across all three documents

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

Price matching Does the invoiced price match the agreed price? Rules Engine Vendor

Numerical comparison with defined tolerance threshold

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

Discrepancy assessment Is the discrepancy within tolerance? Rules Engine

Threshold rule per product group and vendor

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.

Approval routing Who must approve the invoice? Rules Engine

Approval matrix by amount and discrepancy level

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.

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

  • ERP system with procurement module (purchase orders)
  • Electronic goods receipt posting
  • Defined tolerance thresholds per product group
  • Approval matrix by amount brackets

Governance Notes

GoBD-compliant §203 StGB-compliant

GoBD relevance: high - the matching process is the central verification instance between ordering and payment. Fully rule-based (0H/6R/0A) - no AI discretion, no human discretion in the standard case. This makes the agent the ideal starting point for organisations that must demonstrate GoBD compliance. The approval matrix implements the four-eyes principle required in many internal control systems.

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

Process Documentation Contribution

The Three-Way Matching Agent documents for every match: which purchase order and goods receipt were assigned, the exact quantities and prices from all three documents, the calculated discrepancy, the applied tolerance threshold and the routing result. During a tax audit, it is fully traceable whether every invoice was correctly verified.

Infrastructure Contribution

The Three-Way Matching Agent builds the matching pattern reused across the entire Finance organisation. The tolerance threshold logic is used by the Payment Run Agent and Cash Application Agent. The approval routing pattern is the template for all workflow agents that require amount-based approvals.

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

What happens with partial deliveries?

The agent recognises partial deliveries automatically and matches delivered against ordered quantities. Invoices for partial deliveries are checked against the open remainder of the order line. Multiple goods receipts per order are accumulated.

How are price discrepancies handled?

The tolerance threshold is configurable per product group and vendor. Discrepancies within tolerance are approved automatically. Discrepancies outside tolerance are routed to the responsible buyer - with the exact difference and the reference to the agreed price.

Is AI needed for three-way matching?

No. The agent is fully rule-based - all six decision steps are deterministic. That is precisely what makes it the ideal starting point: zero AI risk, maximum traceability, immediate GoBD compliance. AI components only come into play with downstream agents.

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.