What is a reconciliation API?
A reconciliation API is a software interface that automates the matching of financial transaction records across multiple systems, such as payment processors, bank accounts, and internal ledgers, and flags discrepancies that cannot be matched automatically. Rather than downloading bank statement files and comparing them manually, teams integrate the API once and the matching runs continuously in the background.
The API sits between financial data sources and accounting or operations systems. It pulls data, normalizes it into a consistent format, applies matching rules, and returns a structured output showing what matched, what did not, and why.
How does a reconciliation API work?
A reconciliation API typically operates in four stages.
Data ingestion: The API connects to financial data sources and pulls transaction records automatically. Sources typically include bank accounts via BAI2 files or bank APIs, payment processors via their reporting APIs, internal payment ledgers or ERPs, and clearing accounts. Most reconciliation APIs support multiple data formats including JSON, XML, and CSV, and normalize them into a common structure before matching begins.
Matching: The matching engine compares records across data sources using configurable rules. Matching criteria typically include transaction amount, date, reference number, and counterparty. The engine handles different matching patterns:
- One-to-one: a single internal record matches a single bank transaction
- One-to-many: a single settlement amount matches multiple individual transactions in the internal ledger (common with batch settlements from payment processors)
- Many-to-one: multiple internal records aggregate into a single external transaction
Exception handling: Transactions that cannot be matched automatically are flagged as exceptions with a reason. Reasons typically include unmatched transactions, amount mismatches, timing differences, and missing reference data. Exceptions are returned via the API response or surfaced in a dashboard for manual review.
Reporting and audit trail: Once matching is complete, the API generates a reconciliation report showing matched pairs, unmatched items, and exception details. Every match and exception is logged with a timestamp, creating the audit trail required for financial close and regulatory compliance.
What is the difference between a reconciliation API and a payment API?
These two API types are complementary but serve completely different functions.
A payment API initiates transactions. It sends payment instructions to a bank or payment network and returns a confirmation that the instruction was accepted.
A reconciliation API verifies transactions that have already occurred. It pulls records of completed transactions from multiple systems and confirms they are consistently recorded across all of them.
A payment API moves money. A reconciliation API confirms the money moved correctly and is accurately reflected everywhere it should be.
What data sources does a reconciliation API connect to?
The value of a reconciliation API depends on how many data sources it can ingest. A well-integrated reconciliation API typically connects to:
- Banks: Statement files in BAI2 format or direct bank APIs providing intraday transaction data
- Payment processors: Settlement reports from card acquirers, ACH operators, and instant payment networks
- Internal ledger or ERP: The company's own transaction records, which the external data is matched against
- Wire transfer and SWIFT systems: For high-value payment reconciliation across correspondent banking chains
The more sources a reconciliation API can ingest natively, the less data normalization work falls to the engineering team building on top of it.
Why do fintechs use reconciliation APIs?
Manual reconciliation does not scale. A platform processing thousands of ACH and wire transactions per day across multiple banks and processors cannot match records manually. At high volumes, the operational cost of manual reconciliation outweighs the cost of automation, and the error rate makes financial close unreliable.
Reconciliation APIs solve specific problems that emerge at scale:
- Batch settlement mismatch: Payment processors typically settle multiple transactions as a single net deposit. Matching that deposit back to individual transactions requires one-to-many matching logic that manual processes cannot handle reliably at volume
- Multi-rail complexity: Transactions on different rails settle at different speeds and in different formats. A reconciliation API normalizes all of them into a single matching process regardless of which network carried the payment
- Exception velocity: At high volumes, even a 1% exception rate generates hundreds of items per day. Automating the triage of exceptions and routing only genuine discrepancies to human review is what makes the process manageable
The output of a reconciliation API feeds directly into payment reconciliation, bank reconciliation, and transaction reconciliation workflows, providing the structured, matched dataset that finance and treasury teams need for financial close.
