Was the PSP settlement correct?
Validate eligible transactions, refunds, fees, FX adjustments, and expected net settlement.
- Exact and composite matching
- Batch and net settlement support
- Fee and FX difference detection
Payment reconciliation, made explainable
Traqvera Reconcile connects internal transactions, PSP settlements, and bank receipts—then verifies what arrived, when it became available, and why any amount or timing difference occurred.
Built for finance and payment operations teams.
Illustrative product data
Trace. Verify. Resolve.
A clearer financial truth
Payment reconciliation is the process of comparing internal transaction records, PSP settlement reports, and bank entries to prove that the expected amount was settled and became available within the expected time window.
Traqvera turns that process into an operational workflow: it traces the complete money movement, verifies amount and timing at each handoff, investigates exceptions, and preserves the evidence behind every resolution.
Traqvera Reconcile
We do not compare gross transaction value directly with a bank deposit and call the gap an exception. Traqvera separates what the PSP should settle from what and when the bank should receive.
Validate eligible transactions, refunds, fees, FX adjustments, and expected net settlement.
Compare the expected amount and settlement window with the actual bank receipt, keeping amount and timing status separate.
The reconciliation workflow
Six deliberate steps. Every transition visible. Every decision attributable.
Bring together internal transactions, PSP settlement reports, and bank statements.
→Map every source into one financial model and stop bad data before it becomes a false exception.
→Run deterministic matching across transaction-to-settlement and settlement-to-bank stages.
→Collect evidence and use AI to explain the most likely cause of each meaningful difference.
→Approve, correct, reject, or escalate with a structured decision and complete audit trail.
→Sign off a balanced reconciliation and create an immutable report snapshot.
→AI, with financial boundaries
Deterministic rules match records and calculate amount, expected timing, and exposure. AI works where judgment helps: organizing evidence, explaining likely root causes, and recommending the next action.
Calculations stay deterministic. The model never becomes the financial source of truth.
Every claim cites evidence. Records, rules, calculations, and source health remain visible.
People own the outcome. Approve, reject, correct, or escalate before sign-off.
Contracted PSP fees and other legitimate deductions were verified first. The remaining SGD 25.00 residual is supported by the bank entry and fee evidence.
Value proof
Surface confirmed fee, FX, bank-fee, and short-settlement differences instead of treating every unknown as loss.
Attribute delay across PSPs, banks, rails, and settlement legs—while separating money movement from evidence movement.
Separate evidence-pending value from confirmed exposure, so missing evidence is never reported as missing money.
Illustrative product data · Every result is classified as verified, unverified, or confirmed exposure.
Payment reconciliation resources
Practical guides for separating PSP settlement correctness from final bank receipt—and proving every conclusion with evidence.
Explore all resources →A practical guide to reconciling internal payment transactions with PSP settlement reports, including fees, FX, batches, exceptions, and evidence.
Read guide →Settlement control guideHow to reconcile PSP settlements to bank receipts while keeping amount status, timing status, source health, delay, and exposure separate.
Read guide →Commercial assessment guideWhat a Settlement Leakage Audit examines, the data it requires, the controls it applies, and the evidence-based outputs it should deliver in 7–14 days.
Read guide →Financial control guideHow a settlement waterfall separates legitimate deductions, expected net, actual bank receipt, unexplained residual, and confirmed leakage.
Read guide →Questions, answered clearly
What finance, payment operations, and audit teams should know before trusting an automated reconciliation workflow.
No. Traqvera first explains legitimate components such as refunds, chargebacks, PSP fees, FX costs, bank fees, tax, reserves, and adjustments. The amount left after expected net is compared with the bank receipt is an unexplained residual. It becomes confirmed leakage only when evidence supports that conclusion.
No. Amount and timing are evaluated separately. A settlement can match in full and still be late against its expected settlement window, creating time-based exposure even when the amount difference is zero.
Payment reconciliation compares internal transaction records, payment service provider settlement reports, and bank statement entries to verify that expected funds were settled and received—and to explain any difference.
It first compares internal transactions with PSP settlements, then compares expected PSP payouts with actual bank receipts. Keeping those stages separate shows exactly where a difference occurred.
No. Deterministic rules own matching, calculations, and financial state changes. AI organizes evidence, explains likely root causes, and recommends a resolution for human review.
Yes. The initial product is designed for internal transaction, PSP settlement, and bank statement CSV files. Field mapping and data-quality checks happen before a reconciliation run starts.
No. Traqvera Reconcile is an operational control layer. It verifies movement between transaction, settlement, and bank records without pretending to be the accounting system of record.
No. Traqvera observes, verifies, investigates, and documents settlement outcomes. It does not execute payments, move liquidity, trade FX, draw credit, or operate custody or treasury accounts.
Settlement Leakage Audit
Start with a free 30–45 minute fit assessment. If the flow is suitable, a scoped 7–14 day audit uses your internal transactions, PSP settlements, and bank statements to quantify leakage, delay, exposure, and manual workload before implementation.
Create an account