Fintech Engineering

Floats can't represent many decimal values exactly, causing rounding errors. Store amounts as integers in the smallest unit (e.g. paisa/cents) or use a decimal type.

Repeating the same request has the same effect as doing it once. With idempotency keys, network retries won't charge a customer twice.

Every transaction records equal debits and credits across at least two accounts, so the books always balance and every movement is traceable.

Atomicity (all or nothing), Consistency (valid state), Isolation (concurrent transactions don't interfere) and Durability (committed data survives failures).

Use database transactions with row-level locks or optimistic concurrency on the balance, check funds inside the transaction, and use idempotency keys for requests.

Comparing your internal records with external statements (bank, gateway, MFS) to find and resolve mismatches such as missing or duplicate transactions.

Know Your Customer verifies identity; Anti-Money Laundering monitors transactions for suspicious patterns and reports them to authorities as required by law.

Don't assume failure. Mark it pending, query the gateway status or wait for a webhook, and only then confirm or reverse — never blindly retry without idempotency.

The Payment Card Industry Data Security Standard — security requirements for any system that stores, processes or transmits card data. Tokenisation reduces scope.

A way to manage distributed transactions as a series of local transactions with compensating actions to undo steps if a later step fails.