Chapter 6
The most important fraud correction did not happen in a courtroom.
It happened in payroll.
Romano Distribution rebuilt its temporary-labor process.
Not glamorous.
Painful.
Every site had to reconcile:
scheduled headcount.
actual gate entry.
time worked.
vendor invoice.
exceptions.
Warehouse managers complained.
“It adds twenty minutes a week.”
Priya answered:
“You lost seven hundred thousand dollars.”
Complaints decreased.
Then workers raised another issue.
Gate entry is not work time.
A person can enter early.
leave late.
forget badge.
work offsite.
Correct.
So access logs could corroborate but not replace timekeeping.
The system used multiple sources.
Good controls do not turn one imperfect record into absolute truth.
That lesson came directly from M-01.
The access log said Bianca’s card opened the kitchen.
True at credential level.
False if interpreted as:
Bianca physically opened the door.
Records need context.
Then Crestline’s new leadership wanted to preserve the Romano contract.
They proposed:
full transparency portal.
worker verification.
independent audits.
Dante wanted to cancel anyway.
The board asked:
“Why?”
“Trust.”
One director said:
“Trust in whom?”
Alice gone.
Gabe gone.
Santini’s holding stake under forced sale.
new management.
If every company remained permanently defined by old executives, reform had no endpoint.
Romano kept a limited contract for six months under monitor.
Some people criticized.
At six months:
billing clean.
worker complaints lower.
exceptions documented.
Contract renewed at reduced scope.
Not forgiveness.
Performance.
Then Rafael returned from suspension.
Different role:
network planning.
No vendor approvals.
No staffing contracts.
Employees whispered:
family privilege.
Maybe partly.
Dante did not pretend otherwise.
He asked the board to document why Rafael remained.
The answer:
evidence did not support knowing fraud.
he repaid conflicted compensation.
he cooperated.
his operational performance outside vendor oversight was strong.
Could another employee have received same?
The board believed yes.
Whether everyone agreed remained open.
Then Rafael apologized to Bianca.
Not for attack.
For warning Gabe.
“I thought he’d clean it up.”
Bianca said:
“You wanted the problem to disappear before Dante saw it.”
“Yes.”
“So did I.”
They looked at each other.
Different methods.
same control instinct.
Then Rafael said:
“I hated that you suspected me.”
“I did suspect you.”
“I know.”
“You gave me reasons.”
“I know.”
No fake friendship.
Then Bianca’s professional suspension ended.
She returned to work.
Her first assignment involved a family-owned manufacturer with suspicious vendor rebates.
Her supervisor asked:
“What do you do first?”
Bianca smiled bitterly.
“Define authority.”
Then:
“Preserve data through counsel.
notify proper decision-makers.
separate conflicts.
do not steal anyone’s access card.”
Her supervisor almost laughed.
Good.
Then she found a real rebate issue.
A sales director had accepted travel from vendor.
Disclosed late.
not criminal.
company corrected.
Bianca did not build secret spreadsheet at midnight.
She opened formal case.
Progress measured in procedure.
Then Dante began seeing someone new? Too soon perhaps. At two years maybe later. For now none.
He spent more time with family and work.
People tried to set him up.
He refused.
Not because Bianca was “the one.”
Because he was tired.
Then the mansion.
Security upgrades settled into background.
At first every service door beep made Dante look.
Then not.
Sofia’s new master? No universal.
She carried two role badges:
house management.
event cash room.
One day one failed.
Old panic flashed.
Then system showed expired certificate.
IT renewed.
No conspiracy.
Good.
Then a wedding event for another Romano cousin was held at mansion.
Sofia almost declined managing.
Then chose yes.
She walked kitchen route.
checked exits.
approved vendors.
no hidden tests.
During dinner, a server asked:
“Is this where people shot at you?”
Sofia said:
“Yes.”
Then:
“Table six needs water.”
Conversation moved.
She refused to make her trauma a training story for every new employee.
Then Paolo Santini’s criminal case ended in a plea.
He admitted:
conspiracy to steal business records.
obstruction.
participation in staffing fraud through Crestline oversight failures? Let us be precise: he admitted knowledge of inflated billing for some period and efforts to conceal.
He did not admit ordering gunfire.
Evidence did not establish that.
Sentence significant but not life.
Financial penalties.
restitution.
business restrictions.
His freight company continued under new leadership after creditors and board changes.
No empire explosion.
Then Varga received longer sentence due armed intrusion and assault.
Gabe received lesser but substantial sentence for access facilitation and fraud cooperation.
Alice Doran civil settlement and industry ban? Maybe barred from fiduciary role for certain period by settlement. Fine.
Then Sofia read sentencing summaries.
She asked Dante once:
“Does it feel finished?”
He thought.
“No.”
“Why?”
“Because finished sounds like I should stop remembering.”
Sofia shook head.
“Finished can mean no next action.”
That was useful.
The cases were mostly finished.
Memory was not an action item.
Then Dante met Bianca again at a conference reception by accident.
She stood near the exit.
For a second both froze.
Then:
“Hi.”
“Hi.”
“How are you?”
“Better.”
“You?”
“Better.”
No tears.
No legal talk.
Then Bianca said:
“I heard Crestline contract survived.”
“Yes.”
“I’m surprised.”
“So was I.”
“Was it right?”
“So far.”
She nodded.
Then:
“I would have fired them all.”
“I know.”
They smiled.
That was why process mattered.
People wounded by a system should not always be the only people deciding its future.
Then Bianca asked:
“Do you regret canceling wedding?”
Dante took time.
“No.”
Her face tightened.
He added:
“I regret why.”
She nodded.
Fair.
Then she said:
“I don’t regret finding the fraud.”
“You shouldn’t.”
“I regret how.”
“Yes.”
No contradiction.
They parted.
Dante walked back into reception.
He did not look over his shoulder.
The six-month Crestline review also looked at who benefited from the old system besides executives.
Some warehouse managers liked placeholder staffing because it guaranteed extra bodies during peak weeks.
When the fraud controls removed that cushion, staffing became tighter.
Missed shifts rose briefly.
Managers complained:
“The old system worked better.”
Priya answered:
“It worked by hiding cost.”
So Romano negotiated a legitimate standby model.
Crestline could charge a clearly labeled availability fee for workers reserved but not used.
No fake names.
No phantom hours.
The cost appeared as cost.
Managers hated seeing it.
Finance preferred.
Truth often looks more expensive when hidden subsidies disappear.
Then workers liked another change:
cancellation pay.
If Romano canceled a shift after workers had already traveled, they received a minimum payment under new contract.
That was not required by every old agreement.
Dante approved because the prior system had shifted uncertainty downward.
If the company wanted flexible labor, it should pay for some flexibility.
The reform became more balanced than the anti-fraud project he first imagined.
Then Sofia reviewed event staffing at the mansion.
A caterer proposed:
“ten guaranteed, bill twelve if needed.”
She said:
“No. Write ten guaranteed plus two on-call.”
The caterer sighed.
“Your invoices are going to look ugly.”
“Ugly is fine.”
That line spread.
Then an internal audit found no fraud but did find managers manually editing worker names to fix spelling after invoices posted.
That broke traceability.
New process used correction records rather than overwriting.
Again:
do not make history disappear just because the corrected version looks cleaner.
Then Dante asked Priya whether they were becoming obsessed with records.
“Maybe.”
He laughed.
“What’s the test?”
“Can people still do work?”
“Yes.”
“Can they correct mistakes?”
“Yes.”
“Can they understand why fields exist?”
“Mostly.”
“Then you’re probably okay.”
Good.
The controls were not meant to become religion.
They were meant to make ordinary work describe itself honestly.
The new labor portal produced one more lesson.
Workers could now flag missing or incorrect shifts, but the first version required them to create accounts with personal email addresses.
Some agency workers did not want that.
Others changed phones often.
Sofia asked:
“What is the minimum we need?”
The developers changed the portal.
Workers could use agency-issued ID plus one-time verification at the site.
No unnecessary personal account.
Then a manager complained the anonymous-feeling system made fraud easier.
Priya answered:
“Privacy is not the enemy of verification. Bad design is.”
The revised portal still tracked each worker token uniquely.
It simply stopped collecting extra information because someone thought more data always meant more security.
Then Romano’s internal auditors tested the new process with fabricated duplicate profiles.
The system caught four of five.
One slipped through because the same worker token had been entered with a transposed digit.
That was not acceptable.
They added a checksum validation.
Retested.
All five caught.
No press release.
Good controls improve because somebody tries to break them before a criminal does.
Then Dante visited one warehouse during a night shift.
No executives.
He asked a temp worker whether the new system was annoying.
“Yes.”
“Why?”
“Too many scans.”
“How many?”
Three:
arrival.
assignment confirmation.
departure.
The worker said arrival and assignment happened thirty seconds apart.
Redundant.
Dante brought it to audit team.
They combined where safe.
Security should not make people perform bureaucracy merely to prove the system exists.
That was another overcorrection avoided.
One manager asked whether the new system could ever become as invisible as the old one.
Sofia answered:
“Hopefully.”
May you like
A good control should be noticed most when it fails, not every time someone does ordinary work.
That became the standard.
