Automated NDIS payment reconciliation is software matching remittance claim lines back to the original invoices by claim reference, rather than a person tracing each line manually. It flags exceptions — rejected lines, price-cap adjustments, or partial payments — for a human decision, while approved lines that match exactly are reconciled without manual intervention. The practical benefit is speed and consistency at volume, not the elimination of judgment calls on exceptions.
How does automated reconciliation work?
- A remittance is imported (or received via API) containing claim lines referencing participant numbers and claim IDs.
- The software matches each claim line to its originating invoice using the claim reference.
- Lines that match exactly (claimed amount = paid amount) are reconciled automatically.
- Lines that don’t match — rejections, price caps, partial approvals — are flagged as exceptions for review.
- Once every exception has a decision (resubmit, credit, write off), the batch is confirmed as fully reconciled.
What’s the difference between manual and automated reconciliation?
| Manual reconciliation | Automated reconciliation | |
| Matching method | Person traces claim lines to invoices by eye | Software matches by claim reference automatically |
| Speed at volume | Degrades as claim-line count grows | Scales with volume |
| Consistency | Depends on the person doing it, and whether they’re rushed | Same matching logic applied every time |
| Handling exceptions | Risk of silently netting off a rejected line | Exceptions are surfaced and require an explicit decision |
| Best suited to | Low claim volume, few participants | Providers with meaningful claim-line volume per remittance |
Does automation remove the need for manual review entirely?
No — automation handles the deterministic matching (claim reference to invoice, amounts that tie out exactly) but exceptions still need a person to decide the outcome. What changes is where staff time goes: instead of manually matching every line, staff review only the lines the system couldn’t reconcile automatically, which is a smaller and more targeted set of decisions.
