Putting wire confirmation inside the closing platform
The verification call exists because there is no better channel. It costs the officer forty-five minutes, it is defeated by an attacker who phones the buyer first, and afterwards nobody can prove what was said.
How much officer time does wire verification actually cost?
Roughly forty minutes per file, on a control that does not reliably work. Inside a closing platform such as Qualia the same confirmation is an API call, a link to the buyer and a receipt on the file — under two minutes of officer time, and it produces evidence the call never did.
- The verification call is a workaround for the absence of an authenticated channel to the buyer.
- Replacing it with a signed confirmation cuts officer time substantially and produces evidence the call never did.
- The integration is small — issue a statement, send a link, store the receipt on the file — and fits inside existing closing workflows.
Part of Title, escrow and closing wires
What the call actually costs
| Step | Typical time |
|---|---|
| First attempt, no answer | 5 min including voicemail |
| Second attempt, callback scheduling | 10 min |
| Successful call, reading details aloud | 12 min |
| Reading back and correcting | 8 min |
| Noting the call in the file | 5 min |
| Repeat for a second buyer on the file | varies |
Forty minutes is a conservative estimate for a file that goes reasonably well. Across a few hundred closings a year that is a meaningful portion of an officer's capacity spent on a control that does not reliably work.
Why the call is weak as a control
- Direction. If the officer calls the buyer, they may be calling a number from a compromised thread. If the buyer calls in, they may be calling a number from a fraudulent message.
- Audio channel. Reading a routing and account number aloud invites transcription errors, and voice is increasingly weak as an identity signal.
- No artefact. The record is a note saying a call happened. It does not establish what was conveyed or what the buyer subsequently acted on.
- Timing. The call confirms instructions at one moment. The buyer wires later, possibly after receiving something else.
The replacement flow
1. Officer completes the file. Disbursement details are already
in the closing platform.
2. Officer clicks "Request wire confirmation".
→ platform issues a statement containing the instructions,
countersigned by the agency's key
→ a link is sent to the buyer by SMS and email
3. Buyer opens the link on their phone. The instructions render
from the signed statement — not from the message body.
4. Buyer confirms with Face ID or Touch ID.
5. Receipt lands on the file. Officer sees confirmation and the
exact time. Lender can be given a copy.
Officer time: under two minutes.Step three is the part that matters technically. The rendering comes from the signed statement, so a buyer who opens a fraudulent link sees no valid statement at all rather than a convincing forgery.
What the integration requires
Deliberately small, because closing platforms are not systems anyone wants to modify heavily.
- An API call at the point the officer requests confirmation, passing the disbursement fields already in the file
- A webhook or poll to receive the receipt when the buyer signs
- A field on the file to store the receipt and surface its status
- A view for the officer showing confirmed, pending or expired
No change to the closing workflow, the document set, or the disbursement process. The confirmation slots in where the call was.
Handling buyers who cannot or will not
A minority will not have a device with a platform authenticator, or will not complete the flow. The process needs a path for them that does not become the default.
| Situation | Path |
|---|---|
| Modern phone, completes normally | Signed receipt — the standard case |
| Older device without platform authenticator | One-time code to a number verified earlier in the file, plus officer call |
| Refuses or cannot complete | Officer call, documented as an exception on the file |
| Multiple buyers | Each signs; the file shows which confirmations are outstanding |
Treating the fallback as an exception rather than an option is what keeps it rare. If it appears as an equal choice, a meaningful fraction will take it and the control erodes.
What the agency gains beyond time
Three things, in increasing order of value.
- Capacity — officer hours returned to work that requires judgement.
- Timing — a mismatch surfaces before the wire rather than at closing, which is when recovery is still possible.
- Position — in a negligence claim, the agency can produce what it issued and what the buyer signed, rather than a note that a call took place.
The third is the one underwriters will eventually price, and it is the argument that carries this past an operations conversation into a risk one.
What the call costs, itemised
| Step | Typical time |
|---|---|
| First attempt, no answer, voicemail | 5 min |
| Callback scheduling | 10 min |
| Reading details aloud | 12 min |
| Reading back and correcting | 8 min |
| Noting the call in the file | 5 min |
| Second buyer on the file | repeat |
Across a few hundred closings a year that is a meaningful share of an officer's capacity spent on a control with four known failure modes: the number may come from a compromised thread in either direction, account numbers read aloud invite transcription error, the record is a note that a call happened, and the buyer wires later — possibly after receiving something else.
Objections and honest limits
“Some buyers will not complete it.” A minority will not, and they need a documented exception path: a one-time code to a number verified earlier in the file, plus an officer call. Keeping it an exception rather than an equal option is what preserves the control.
“This is an operations change, not a risk one.” It is both, and the risk half is what underwriters will eventually price. The agency can produce what it issued and what the buyer signed, rather than a note that a call took place.
Integrating with a closing platform
- API call at the confirmation request. Passing the disbursement fields already in the file.
- Webhook to receive the receipt. When the buyer signs.
- A field on the file to store it. With status: confirmed, pending or expired.
- An officer view. Showing which confirmations are outstanding across the book.
Terms used here
- Disbursement details
- The bank account, routing number and beneficiary the closing funds are sent to. The field attackers substitute.
- Out-of-band verification
- Confirming instructions through a channel independent of the one that delivered them. Only as good as how the second channel was obtained.
- Exception path
- The documented route for buyers who cannot complete the standard flow. It stays rare only if it is treated as an exception rather than an option.
Frequently asked questions
Why is the verification call weak? The number may come from a compromised thread in either direction, reading account numbers aloud invites error, and the record is a note that a call happened rather than what was conveyed.
What if the buyer has no compatible device? A documented exception path — one-time code to a previously verified number plus an officer call. Keeping it an exception rather than an option is what preserves the control.
How much integration work is this? An API call when the officer requests confirmation, a webhook to receive the receipt, and a field on the file. The closing workflow is unchanged.
Does the buyer see the instructions in the message? No. The link opens a rendering of the signed statement. A fraudulent link produces no valid statement rather than a convincing copy.
Where this fits in Manav
Manav is the confirmation step inside the closing software: one API call out, one receipt back, stored on the file. The officer's forty minutes become two, and the agency gains an artefact it can produce in a claim.