Skip to content

Reconciliation

Reconcile different sources.Understand every difference.

Match banks, ERPs, card acquirers and documents using your own business rules, with AI support. What matches is reconciled automatically; what doesn't goes to review, with the source of the data and the history of every decision.

  1. ERP · Accounts receivable

    Invoice 1042R$12,450.00Aurora Serviços · Inv. 8842Due 11/04

    Bank · Statement

    Pix receivedR$12,450.00AURORA SERVICOS · NF884211/04
    • Met:DocumentInvoice 8842 in the Pix description
    • Met:AmountSame on both sides
    • Met:DateCredited on the due date
    Reconciled by rule
  2. ERP · Accounts receivable

    Invoice 1063R$4,300.00Cedro Comércio · Inv. 8860Due 10/30

    Bank · Statement

    Pix receivedR$4,214.00CEDRO COM LTDA11/06
    • Met:CounterpartyCustomer's CNPJ matches
    • Not met:AmountR$86.00 short, no discount in the rule
    • Informational:Date7 days after the due date
    Difference to review

Illustrative data.

The routine

The work piles up.When every check starts from scratch.

Statements, open items, sales and documents arrive at different times and in different formats. Matched by hand, differences surface late, open items scatter across spreadsheets and the close waits.

  1. Every day, 9 AM

    The same check, again.

    Export each bank's statement, filter it, hunt for the invoice number in the description. Tomorrow, all over again.

  2. Day 3 of the close

    The difference shows up late.

    The overcharged fee or the credit that never came only surfaces when the month's balance won't tie out.

  3. Audit, months later

    Nobody remembers why.

    Who accepted that difference, and on what basis? The answer lives in a highlighted cell or in someone's memory.

Bank and open item reconciliation

Does the ERP match the bank?Item by item, with the criteria in view.

Payables and receivables matched to bank transactions. Amount, date and reference are checked to identify settlements, partial payments and what needs review.

  • ERP open items and entries compared with bank debits and credits.
  • Full or partial payments and receipts.
  • Interest, late fees and discounts recognized according to the configured rules.
  • Open items with no transaction and transactions with no open item set aside for review.
  • Clearing of the reconciled item sent to the ERP, when the integration allows it and the company authorizes it.

Transactions through Open Banking, for European and Brazilian banks, and direct bank integrations.Bank and crypto connectivity

Criteria · Invoice 1042

  • Met:DocumentInvoice 8842 appears in the Pix description.
  • Met:AmountR$12,450.00 on both sides.
  • Met:DateCredited on the due date.

Result

Reconciled by rule

Clearing of the open item sent to the ERP.

Illustrative data. Pick an example to see its criteria.

Sales and receivables

From the sale to the bank credit.Every step checked.

Sales, card acquirer data and amounts received, matched to one another. Fees, installments and payouts checked to separate what was expected from what actually happened.

  • Recorded sale × transaction captured by the acquirer.
  • Expected fee × fee charged.
  • Installments and expected payout dates.
  • Early payouts, cancellations and chargebacks.
  • Expected net amount × bank credit.

Sales and fees through direct integrations with card acquirers; how each one connects is set during rollout.

1Check the sale

Does the recorded sale exist at the acquirer, for the same amount?

Sales system

Order 48213R$1,000.00Credit card, single payment · 11/03

Card acquirer

Captured transactionR$1,000.00Authorization 004817 · 11/03
  • Met:AuthorizationSame code on both records.
  • Met:Gross amountR$1,000.00 on both sides.
  • Met:DateSale and capture on the same day.
Sale verified

2Check the settlement

Did what should reach the bank actually arrive?

Gross saleR$1,000.00
Minus Contracted fee (3.00%)R$30.00
Equals Expected netR$970.00
Bank credit · 12/03R$965.00
Difference R$5.00 shortTo investigate

Illustrative example, with amounts consistent across steps.

Other applications

Other relationships in your operation.Each with its own sources and criteria.

Documents, obligations and accounting balances checked with the same logic: defined sources, explicit criteria and differences set aside.

Rules and AI

Your business rules.To decide what makes a pair.

Identifiers, documents, amounts and dates combined into rule sets, one per reconciliation type. When there is no exact match, AI reads the history and context and suggests a link for your team to review.

  1. 01 · Records compared

    Order 48213R$89.00Pix · 11/05withPix receivedR$89.00Order ID in the transaction · 11/05
  2. 02 · Criteria met

    • Met:IdentifierThe order ID is in the transaction.
    • Met:AmountR$89.00 on both sides.
    • Met:DateSame day.
  3. 03 · Result

    Reconciled by rule

    1 record × 1 record

Illustrative data.

  • Reconciled by rule

    Every criterion in the rule was met. It goes through without review.

  • Suggested by AI

    With no exact pair, AI proposed a link based on history and context. Your team confirms it.

  • No match

    No record meets the criteria. It goes to exceptions.

  • Same amount isn't enough: two R$89.00 Pix payments on the same day only become a pair when an identifier, document or counterparty confirms it.
  • Categorizing says what a transaction is. Reconciling shows which record it matches.

Exceptions

Every exception with context.So you resolve it, not hunt for it.

The records involved side by side, the difference calculated and the condition identified. The item stays open until your team decides.

Open items · 3

Invoice 1063 · Cedro Comércio

Amount difference
ERP
Invoice 1063 · Inv. 8860 · due 10/30
Bank
Pix received from CEDRO COM LTDA · 11/06
Expected
R$4,300.00
Actual
R$4,214.00
Difference
− R$86.00
Condition identified
Amount received is lower than the invoice, and no rule allows a discount for this customer.
Evidence
Invoice 1063 in the ERPPix of 11/06, via Brazil's Open Finance
Decision
Pending. No clearing in the ERP until your team decides.
Next step
Confirm with the customer whether a discount was negotiated or the balance will be paid.

Illustrative data. Pick an open item to see the details.

Automatic

Goes through automatically

Pairs that meet the rules, including interest, late fees, discounts and installments they allow for. The reconciled item is cleared in the ERP when the integration allows it and the company authorizes it.

Review

Needs your team

Links suggested by AI, differences outside the rules, records with no match and possible duplicates. They stay open, with no clearing, until someone decides.

Traceability and close

From the source record.To evidence for the close.

Sources, criteria and decisions stay attached to each reconciliation. It's clear what was reconciled, what is still open and what the ERP has already confirmed.

History

Invoice 1042 · Aurora Serviços · R$12,450.00

  1. Data received

    Received

    Pix of 11/04 via Brazil's Open Finance; invoice 1042 read from the ERP.

  2. Rule applied

    Applied

    Accounts receivable: document, amount and date.

  3. Result

    Reconciled by rule

    All three criteria met.

  4. Review

    Not applicable

    Not needed: the rule resolved the pair. An AI suggestion or a difference would come through here.

  5. Clearing requested

    Sent

    Clearing of the open item sent to the ERP.

  6. ERP response

    Confirmed by the ERP

    The ERP returned the invoice as settled.

Illustrative data.

Three different states

  • Reconciliation complete

    The pair exists in Openi, with sources and criteria.

  • Clearing requested

    Sent to the ERP, no response yet.

  • Confirmed by the ERP

    The ERP returned the update. Depends on the integration reading that response.

For the close, reconciliation delivers the pairs with evidence and the list of what is still open. The accounting close itself stays in your process and your system.

Cases

Reconciliation in production.Process, sources and results.

NDD

Reconciliation at scale

More than 100 million records reconciled

  1. Data sources
  2. Rule set
  3. Matching
  4. Pairs with a decision
Read the full story
Process reconciled
Records from two sources compared by rule sets, at scale.
Sources
Files, database tables and webhook events.
How it decides
Each pair comes back with a decision and a score. What matches goes through; possible matches and unmatched records go to review.
Result
More than 100 million records reconciled. The period for that total was not disclosed.

Proven results

records reconciled
100M+

FAQ

Before you reconcile.What people ask before signing up.

Didn't find what you were looking for?

Talk to a specialist and get your answer straight from our team.

Talk to a specialist

Bank transactions through Open Banking (European and Brazilian banks) and direct integrations, ERP open items and entries, card acquirer sales and fees, captured invoices, CSV and XLSX files, databases and events your systems send by webhook. The sources for each operation are set during rollout.

Demo

See reconciliationapplied to your operation.

Tell us which data sets your team compares and where the biggest pain points are. We'll demo a flow for that scenario.

  1. You leave your details

    It takes less than a minute.
  2. A specialist gets in touch

    To understand your banks, your ERP and what takes up the most time today.
  3. You see Openi at work

    A reconciliation flow built on the data sets you compare.

Demo environment: details are not stored or sent.