Custom Rules Engine for Reconciliation

Finance teams often need more than a one-size-fits-all matching flow. Different reports, identifiers, amount structures, and settlement patterns require different rules. Cointab's custom rules engine for reconciliation lets teams define how records should be compared, matched, grouped, and reviewed so the workflow reflects the way the business actually operates.

Why finance teams need a custom rules engine

A reconciliation process may look simple on paper, but in practice the logic changes from one workflow to another. A sales report may need to be matched to payment gateway data by order ID and amount. A bank reconciliation may need to compare narration, reference numbers, and net amounts. A marketplace settlement may need to account for commissions, returns, deductions, and partial settlements.

When teams rely on fixed spreadsheet logic, the process becomes hard to maintain and difficult to audit. Custom formulas, VLOOKUPs, and manual checks can work for a single file, but they do not scale well across recurring reconciliations.

A configurable rules engine helps finance teams:

• apply different logic to different reconciliation workflows
• handle one-to-one, one-to-many, many-to-one, and many-to-many matches
• compare identifiers using different methods such as equals, contains, or similar logic
• net amounts where needed
• separate matched, partially matched, unmatched, and skipped records clearly
• keep the setup reusable for future periods

How Cointab's reconciliation rules engine works

Cointab is built around a structured Side A and Side B model.

• Side A contains your records, such as sales, books, ERP exports, order data, or internal ledgers.
• Side B contains external records, such as payment gateway reports, bank statements, marketplace settlements, vendor statements, or delivery partner data.

To define a reconciliation, users typically:

1. create a new reconciliation in a team workspace
2. choose a popular reconciliation or create a custom one
3. upload the required files on both sides
4. map the key fields such as date, amount, and identifier columns
5. add supporting data if needed for lookups, enrichment, or calculations
6. create derived columns where additional logic is required
7. define the matching rules
8. run reconciliation manually or on a schedule
9. review the report and exception buckets

The result is a repeatable workflow that finance teams can run again for future periods without rebuilding the setup from scratch.

What you can define in the rules engine

Cointab supports a flexible reconciliation setup so teams can configure logic that fits their business rules.

Matching logic

Users can configure how records should match across the two sides. This includes:

• one-to-one matching
• one-to-many matching
• many-to-one matching
• many-to-many matching
• partial matching
• contra matching
• net-to-net matching

Identifier rules

Reconciliation can be based on one or more identifiers, such as:

• order ID
• transaction ID
• invoice number
• payment reference
• bank UTR
• settlement ID
• AWB number
• customer code
• vendor code
• SKU

Rules can compare a single identifier field or multiple identifier fields together. This is useful when external reports use different reference formats or when a business needs to match on a combination of columns.

Comparison methods

The engine supports different comparison methods depending on the workflow, including:

• equals
• contains
• similar
• equals subset
• contains subset
• similar subset

These options help teams work with reports that do not always share identical formats.

Amount logic

Not every reconciliation needs a simple amount-to-amount check. Some workflows require amounts to be combined, offset, or adjusted before comparison. Cointab supports structured matching logic so teams can compare records in a way that reflects the underlying business process.

Derived columns

If a workflow needs calculated fields, users can create derived columns on both sides. These can be used as amount fields, identifier fields, lookup fields, or output fields.

Examples include:

• clean order ID
• net amount
• delivered payment amount
• normalized transaction ID
• amount after fee
• refund amount as negative
• combined reference
• clean AWB number

Users can also use AI to generate Excel-style formulas from natural language, which reduces manual formula building while keeping the logic reviewable.

Where custom rules are most useful

A custom rules engine is especially helpful when standard reports need business-specific logic.

Common use cases include:

• sales vs payment gateway reconciliation
• marketplace sales vs settlement reconciliation
• bank statement vs books reconciliation
• vendor ledger vs vendor statement reconciliation
• order report vs COD delivery partner reconciliation
• internal sales report vs multi-PSP reconciliation
• ERP sales vs marketplace settlement reconciliation

These workflows often involve multiple files, inconsistent references, deductions, returns, fees, and delayed settlements. A rules-based approach makes it easier to separate true matches from open exceptions.

How exception handling stays clear

One of the most useful parts of a custom rules engine is the way it surfaces exceptions.

Cointab separates transactions into clear buckets:

• fully matched
• partially matched
• unmatched
• skipped

This helps finance teams focus on the records that need review instead of checking every row manually.

• Fully matched means the records match according to the configured rules.
• Partially matched means the records are related, but the amounts do not fully align.
• Unmatched means the record could not be found on the other side.
• Skipped means the row was not included because of a file issue, missing data, invalid amount, duplicate row, or an exclusion rule.

Open items can then be investigated with filters, manual review, or AI-assisted analysis.

AI support for difficult open items

Cointab uses AI in a conservative, finance-friendly way. It does not replace the rules engine. Instead, it helps when deterministic logic is not enough.

AI can assist with:

• creating derived column formulas
• analyzing open transactions
• identifying possible reasons for mismatches
• suggesting likely next actions
• handling unstructured references or inconsistent descriptions

If the evidence is not strong enough, the record should remain unmatched rather than being forced into a weak match.

Reusable setup for recurring reconciliation

A strong reconciliation rules engine should not need to be rebuilt every month. Once the workflow is configured, teams can reuse it for future periods by uploading the new files and running the same setup again.

This is useful for:

• monthly close processes
• daily payment reconciliation
• recurring marketplace settlement review
• vendor statement checks
• bank reconciliation cycles
• quarter-end or year-end reporting

Cointab also supports missed file upload and report refresh, which is practical when a partner report arrives late and the reconciliation needs to be updated.

Automation and team control

After a custom reconciliation is configured, teams can automate the flow using email, SFTP, or API. That allows data to be received, validated, reconciled, and reported on a recurring schedule without manual rework.

Cointab also supports team workspaces, roles and permissions, audit logs, and downloadable Excel reports. That makes the reconciliation process easier to share across finance, accounting, audit, and operations teams while keeping the workflow visible and reviewable.

Why a custom rules engine matters

For finance teams, the value is not just automation. It is control.

A custom rules engine helps teams know:

• what files were used
• what rules were applied
• what matched
• what remained open
• what was skipped
• what needs review next

That combination of flexibility, transparency, and auditability is what makes reconciliation easier to manage at scale.