$
Premium In
Match expected premium to bank cash using policies, endorsements, broker remittances and bordereaux. Flag unidentified cash and short-pays.
AI reconciliation for insurance
Insurance reconciliation, continuously.
Dream Hannah matches premium, commissions, bordereaux, claims, refunds, payment instructions, and bank activity across carriers, MGAs, TPAs, brokers, and programs—so finance teams see breaks while action is still possible.
Premium InCommissionsBordereauxClaims OutRefunds
DHInsurance Reconciliation
Current period ▾
Premium in$428.6Mcurrent period
Claims out$312.4Mcurrent period
Exceptions186ranked for review
Matching statusLiveevidence attached
Matching trend
Exception breakdown
MatchedAI reviewExceptionOpen
Top exceptions
| Type | Amount | Evidence | Status |
|---|---|---|---|
| Unmatched premium | $41,200 | No policy reference | Review |
| Commission variance | $18,500 | Outside contracted terms | High |
| Duplicate claim | $12,300 | Same claim/payee | High |
Across insurance money movement
Paid right. Paid once. Proven.
Dream Hannah checks what should have happened against what actually happened across both sides of the insurance ledger.
%
Commissions & Bordereaux
Check broker deductions against contracted terms and reconcile line-item bordereaux to remittances and bank activity.
⇢
Claims Out
Prove approved claim → payment instruction → bank debit. Surface duplicates, wrong payees, failed payments and reissues.
⚑
Exceptions & Recovery
Rank the unresolved tail by materiality and risk, preserve the evidence, and track investigations through closure.
The AI engine
Three layers make the system auditable by design.
01
Deterministic rules, orchestrated by AI
Exact and tolerance matches run first. Transparent rules decide the clean cases.
02
AI analysis for the ambiguous tail
Narrative matching handles combined payments, missing references, aliases and messy context.
03
Human-in-the-loop learning
Reviewer decisions teach the engine so recurring broker and provider quirks become easier over time.
Inputs
▤
Bank feedsDaily, multi-currency
▧
Remittances & statementsAny format
◎
Policy & claims extractsSource systems unchanged
Dream Hannah engine
1
Normalize & mapPrograms, aliases, counterparties
2
Rules-based matchingExact + tolerance first
3
AI analysisAmbiguous tail
Outputs
✓
Matched recordsEvidence attached
!
Ranked exceptionsHumans touch flagged items
⌁
Audit trailEvery decision traceable
What Dream Hannah catches
The risk lives in what almost agrees.
| Exception | What it can mean | Illustrative signal | Risk |
|---|---|---|---|
| Unmatched premium | Cash with no matching policy or record | ACH deposit not tied to policy | Medium |
| Short-pay | Payment below expected amount | Commission, tax, fee or partial installment | Medium |
| Commission variance | Broker deduction outside contracted terms | Deduction exceeds contracted rate | High |
| Duplicate claim | Same claim / payee / amount paid twice | Second debit posts against one claim | High |
| Changed bank instructions | Beneficiary details changed before payment | New account just before payout | High |
| Cancelled-policy premium | Cash continues after schedule ends | Remittance after cancellation | Medium |
Fraud & payment integrity
Catch what passed the front door.
A changed payee account or duplicate claim can look legitimate at authorization. Reconciliation checks the payment against the claim, instruction, payee and bank debit that should explain it.
Changed payee detailsDuplicate claimsWrong-account paymentsOrphan debitsFailed / reissued payments
Onboarding
Start with your data. Tune on live workflows.
Dream Hannah is configured around your systems, matching rules, aliases, tolerances, and exception workflows—then refined with your team as real cases are reviewed.
Connect
Bring the evidence together
Bank data, policy and claims extracts, remittances, bordereaux, payment records, and account-specific rules.
Configure
Set the matching logic
Map formats, counterparties, aliases, tolerances, approval paths, and edge cases.
Run live
Review the exceptions
Clean cases clear through rules; ambiguous cases arrive with context and evidence for review.
Learn
Keep the knowledge
Reviewer decisions strengthen the operating model and preserve institutional knowledge.
Close insurance books faster. Detect risk earlier.
See Dream Hannah on your premium and claims workflows.