Continuous Matching
Match bank activity, payment records, remittances, bordereaux, policies, claims, ledgers, and source records as data arrives.
Dream Hannah continuously matches money in and money out to the records behind it—so finance teams close faster, see exceptions while action is still possible, and catch errors and fraud before they become permanent.
| Flow | Amount | Record | Status |
|---|---|---|---|
| Insurance premium | $148,220 | Policy 82714 | Matched |
| Owner distribution | $82,144 | Ledger 5844 | Matched |
| Claim payment | $12,500 | Claim 10983 | AI review |
| Vendor payout | $6,980 | Invoice 8810 | Exception |
Hannah works 24/7 to reconcile the books, match expected vs. actual money movement, investigate exceptions, and keep the evidence attached to every transaction. She helps finance and operations teams close faster, protect every dollar, and spend less time assembling the story behind each payment.

Matching is the starting point. Hannah turns the unmatched tail into a ranked worklist with context, reasoning, and evidence—so people spend time deciding, not assembling.
Match bank activity, payment records, remittances, bordereaux, policies, claims, ledgers, and source records as data arrives.
High-confidence matches clear automatically. Ambiguous items arrive prioritized by risk, value, and uncertainty with reasoning attached.
Surface changed banking details, duplicate payments, missing deposits, wrong payees, short-pays, and other breaks while recovery may still be possible.
Every match and exception carries a traceable evidence chain so close, audit, and investigations start from proof instead of reconstruction.
Dream Hannah combines deterministic matching, AI analysis for the ambiguous tail, and human-in-the-loop learning so the engine improves as your team resolves real exceptions.
Bank feeds, system extracts, files, payment records.
Map formats, counterparties, aliases, references.
Apply exact and tolerance rules across the book.
Use AI for missing or messy context.
Rank exceptions that need human decisions.
Produce matched records, reports, audit trails.
Match premium, commissions, bordereaux, claims, refunds, and bank activity across carriers, MGAs, TPAs, brokers, and programs.
Explore Insurance →Reconcile rent, trust accounts, deposits, vendor payments, refunds, and owner distributions across properties and portfolios.
Explore Property Management →Add reconciliation and exception intelligence to embedded payment flows so platform transactions stay connected to the business events behind them.
See the platform model →Dream Hannah ties the money to the policy, claim, broker, bordereau, remittance, payment instruction, and bank activity that explain it.
| Flow | Expected | Actual | Difference | Status |
|---|---|---|---|---|
| Premium · Policy 20418 | $95,000 | $95,000 | $0 | Matched |
| Broker remittance | $48,500 | $42,500 | $6,000 | Short-pay |
| Claim · CLM-66218 | $12,300 | $24,600 | $12,300 | Duplicate |
| Return premium | $4,200 | $4,200 | $0 | Matched |
Dream Hannah checks whether the money that moved agrees with the resident, property, owner, vendor, deposit, and bank records behind it.
| Flow | Amount | Property | AI insight | Status |
|---|---|---|---|---|
| Rent receipt | $2,450 | Oakview · 12B | Lease matched | Matched |
| Security deposit | $2,450 | Oakview · 12B | Trust posted | Balanced |
| Vendor payment | $8,200 | Pinnacle · Miami | Bank details changed | Review |
| Owner distribution | $19,760 | Sunset Villas | Ledger agrees | Matched |
For software platforms that collect or send money, Dream Hannah can sit behind the payment flow as a matching and exception layer—connecting transaction records back to invoices, customer records, platform events, and systems of record.
| Platform event | Payment | System record | State |
|---|---|---|---|
| Invoice paid | $4,820 | INV-23881 | Matched |
| Supplier payout | $1,220 | PO-1182 | Matched |
| Refund | $642 | Order 92771 | Needs context |
| Duplicate payout | $8,440 | PAY-8812 | Exception |
Fraud can look legitimate at authorization. Reconciliation asks a different question after money moves: did this payment actually agree with the invoice, vendor, lease, policy, claim, beneficiary, bank instruction, and expected cash flow behind it?
You cannot stop every attack at the door, but you can take away the time.Brent Ho-Young · CEO & Co-Founder, Dream Payments
Fraud context: the launch materials cite AFP reporting that 76% of organizations experienced attempted or actual payments fraud in 2025, and FBI reporting more than $3 billion in business-email-compromise losses.
No. Hannah is a reconciliation agent: it ingests financial and operational records, matches expected dollars to actual dollars, ranks exceptions, preserves evidence, and learns from reviewer decisions.
No. Clean exact and tolerance matches are handled with deterministic rules. AI is used to orchestrate rules and reason through ambiguous cases. Exceptions can route to a human reviewer.
No. Hannah is designed to work alongside existing accounting, policy, claims, property-management, and payment systems.
Reviewer decisions teach the engine how your counterparties, references, aliases, tolerances, and recurring edge cases behave. That institutional knowledge stays in the system and helps reduce repeated manual investigation.
Start with a reconciliation assessment to estimate match coverage, identify exception exposure, and see where Dream Hannah can create the fastest impact.