AI Reconciliation: How BEST Keeps Finance in Control

A supplier statement includes an invoice that doesn’t appear in SAP. A customer sends payment advice covering several invoices, but the payment received is less than the total.

AI can help read the information in those documents. Completing the finance process means comparing it with SAP records, identifying what agrees and working through the differences.

BEST supports AI reconciliation by combining focused AI document-reading capabilities with its established reconciliation and clearing processes. The BEST Vendor Recons module supports supplier statement reconciliation, while the BEST Customer Clearing module supports customer payment matching and clearing.

Following these two workflows shows where AI contributes, how BEST uses the information and which decisions remain with finance.

What does AI reconciliation involve?

AI reconciliation means using artificial intelligence to support tasks within a financial reconciliation process. Depending on the solution, those tasks might include interpreting documents, suggesting matches or helping analyse exceptions.

BEST uses ChronoScan and language-model APIs to read varied finance documents and return structured information to the relevant BEST process. For supplier statements and customer payment or remittance advice, this information feeds into reconciliation or clearing in SAP.

The distinction matters when assessing functionality. Document reading, transaction matching, clearing and approval are separate tasks. A process can combine AI with configured matching logic and workflow automation.

Our article on AI document processing for SAP finance teams explains the document preparation in more detail. Here, we’ll follow the information into the finance processes it supports.

How BEST supports AI reconciliation for supplier statements

Supplier statement reconciliation compares a supplier’s record of the account with the organisation’s accounts payable records.

The statement might contain invoices, credit notes and payments. Some entries will agree with SAP. Others may be missing from one side or recorded differently.

The BEST Vendor Recons module automates comparison between supplier statement information and SAP accounts payable data. It supports automatic matching, suggested and manual matches for exceptions, reconciliation reporting and approval using SAP authorisations.

AI document reading supports the information entering that process.

From statement to reconciliation result

Consider an illustrative supplier statement containing several invoices, a credit note and a payment. Most entries correspond with SAP, but three differences remain.

The following example shows the roles of AI, the BEST Vendor Recons module and finance. It isn’t a customer case study.

This gives accounts payable a process for moving from the supplier’s document to a reviewed reconciliation result.

Example: an invoice is missing from SAP

In this illustrative example, the statement includes an invoice for £1,250, but there is no corresponding entry in SAP.

The information may have been read correctly. The next task is to establish why the records differ.

Accounts payable may need to check whether:

• The invoice was received.
• It is awaiting processing or approval.
• It was posted using a different reference.
• It belongs to another account or period.
• The supplier has included it incorrectly.

The BEST Vendor Recons module supports comparison and reporting, while finance investigates the underlying cause. Its suggested and manual matching capabilities can help users address exceptions where the corresponding records can be established.

If the invoice hasn’t been received, the appropriate action may be to request it from the supplier. If it is already being processed, the team needs to confirm its status.

The reconciliation has identified work that needs attention. AI document reading has supplied the statement information; it hasn’t diagnosed why the invoice is missing.

Different discrepancies need different follow-up

Other items on the same statement may require different treatment.

The BEST Vendor Recons module supports sending reconciliation reports to suppliers, giving the team a basis for discussing specific differences.

It also tracks missing statements. This helps distinguish accounts that have been reconciled from those where the required statement hasn’t yet been received.

Approval follows the reconciliation requirements

The BEST Vendor Recons module supports multiple approval levels and criteria using SAP authorisations. An optional payment block can also be used where there is no statement match.

These controls govern how the reconciliation progresses. Finance determines the organisation’s requirements and who is authorised to approve.

AI supplies information to the reconciliation process. Matching logic, SAP authorisations and the configured BEST approval process govern what happens next.

How BEST supports customer payment allocation

Customer payment advice describes how a customer intends a payment to be allocated. It may list many invoices, credits or deductions.

The BEST Customer Clearing module supports matching customer remittance information to open receivables and clearing customer items in SAP. AI document reading helps capture the payment advice information used in that process.

The workflow connects three records:

• The customer’s advice.
• The open items recorded in SAP.
• The payment being allocated.

Agreement between two of these doesn’t necessarily establish agreement across all three.

Example: the customer pays less than the advice total

Consider an illustrative remittance listing invoices totalling £10,000, with a payment received for £9,800. This is a process example rather than a customer result.

AI document reading can help capture the listed references and amounts. The BEST Customer Clearing module supports matching that information with SAP open items.

The £200 difference still needs an explanation.

The customer might have deducted a disputed amount, taken a discount or made a payment error. Finance needs evidence to establish which explanation applies and whether the treatment is acceptable.

SAP’s explanation of document clearing describes how clearing balances an open item or group of items to zero.

That accounting action needs to follow the organisation’s rules. Capturing a deduction on a remittance doesn’t establish that finance has accepted it.

Keep unresolved items visible

The objective is to automate the matching and clearing work that can be completed under the defined process, while giving finance visibility over what remains.

The BEST Customer Clearing module provides visibility of remittance lines that haven’t matched. This helps users investigate outstanding items rather than relying solely on a lump-sum payment balance.

For example, a missing invoice reference may require clarification from the customer. A disputed invoice may need input from the account manager. A payment difference may require supporting documentation before it can be treated appropriately.

Each is a finance task with its own evidence and owner.

How finance keeps control of AI-supported reconciliation

Across these workflows, control depends on being able to understand how incoming information was used and what action followed.

For supplier reconciliation, the BEST Vendor Recons module combines matching, reconciliation reports and approval capabilities. For customer payments, the BEST Customer Clearing module supports matching remittances to open receivables, clearing items and identifying unmatched remittance lines.

The capabilities serve different processes, so they should be assessed against the work each team needs to complete.

When reviewing AI reconciliation in SAP, ask to see:

• A successful result: Which entries matched, and which SAP records do they relate to?
• An unresolved item: What remains outstanding, and what information can the user investigate?
• A manual decision: What can an authorised user change or confirm?
• The subsequent action: Does the result lead to reconciliation approval, customer clearing or further investigation?

These questions help establish how AI input, established automation and finance judgement work together.

For the technical background to BEST’s current capabilities, see SAP AI for finance.

Frequently asked questions

Does BEST use AI to reconcile supplier statements?

BEST uses AI document reading to help prepare supplier statement information. The BEST Vendor Recons module then uses that information in its SAP reconciliation process, supporting matching, reporting, exception handling and approval.

Does BEST use AI to clear customer payments?

AI document reading captures customer payment and remittance advice information. The BEST Customer Clearing module supports the subsequent matching and clearing process in SAP. AI document reading doesn’t independently decide how an unresolved deduction should be treated.

Is automated matching the same as AI reconciliation?

Automated matching can use configured rules and algorithms without AI. AI may support another part of the same workflow, such as reading an incoming document. It’s useful to establish which task each capability performs.

What happens when records don’t agree?

The difference needs to be investigated. Depending on the process, finance may need to locate an invoice, confirm a payment, clarify a deduction or obtain information from a supplier or customer.

See how BEST supports your reconciliation workflow

Bring a supplier statement or customer remittance scenario to a demonstration and explore how the relevant BEST module supports the process, from incoming information to matching, investigation and the appropriate finance action.

Book a demo to see how BEST supports AI-assisted reconciliation and customer clearing in SAP.