The entries were brief.
ROUTINE REVIEW.
REFER TO DISTRICT SUPERVISOR.
LIMIT DISTRIBUTION.
EXECUTIVE NOTICE REQUIRED.
Near the bottom, one entry had been highlighted.
The name was Andre Willis.
Elena recognized it from the earlier complaint involving a man who had recorded officers searching a parked vehicle.
Beside his name was a notation:
HIGH EXPOSURE — POTENTIAL MEDIA DISTRIBUTION.
The recommended action read:
CONTACT EMPLOYER. REVIEW ASSOCIATED RECORDS. MONITOR FUTURE COMPLAINTS.
Elena felt anger rise, but she forced herself to remain focused.
The document was troubling, yet its meaning depended on context. The investigators needed to determine who had created the spreadsheet, whether the recommended actions had actually been carried out, and whether the classification system had been authorized for legitimate purposes or used to interfere with complaints.
She called Priya.
“Do we have a date?”
“The spreadsheet metadata indicates it was created four years ago. The last modification was eleven months later.”
“By whom?”
“The account name is ‘NSTAR-ADMIN.’ That is not a personal identity.”
“Can you identify the user who logged in when it was created?”
“We are working on that. The computer was connected to an internal network at some point, but the access logs are incomplete.”
“Look for backup copies. And compare the document with the versions stored in the department’s official system.”
“Already underway.”
Elena opened her own case file and found the original complaint involving Andre Willis.
The official record contained a short summary and a note stating that no further action was recommended.
There was no reference to employer contact.
There was no indication that Willis had been classified as a potential media risk.
There was no mention of the spreadsheet.
If the archive was authentic, then someone had been maintaining a second record system that tracked complaints according to their potential to create institutional embarrassment.
The question was whether the second system had merely monitored complaints or had been used to manipulate their outcomes.
At 10:12, another message arrived from Priya.
WE FOUND A SECOND FILE. IT CONTAINS DRAFT LANGUAGE FOR CLOSURE MEMOS. SEVERAL SENTENCES MATCH THE FINAL REPORTS IN THE OFFICIAL SYSTEM.
Elena read the message twice.
She replied immediately.
PRESERVE BOTH VERSIONS. COMPARE METADATA, AUTHORS, AND APPROVAL HISTORY. DO NOT ASSUME THE DRAFTS WERE USED UNTIL THE RECORDS ESTABLISH THAT CONNECTION.
Then she called David.
“If the same wording appears in the archive and in the official closure memos, we may be able to establish that the hidden system influenced the formal investigations.”
“That would be significant,” David said. “But we will need to distinguish copied language from evidence of improper direction.”
“Agreed.”
“Also, the search team found a document referring to an external review committee. The committee supposedly evaluated cases involving potential legal exposure.”
“Who served on it?”
“We do not know yet. The document lists departments rather than individuals.”
“Then we need the meeting records.”
“We are looking for them.”
Elena walked to the conference room and wrote three questions on the whiteboard.
WHO CREATED NORTHSTAR?
WHO USED IT?
WHO BENEFITED FROM IT?
She underlined the final question.
Corruption investigations often became consumed by identifying the person who signed a document. But signatures were only one part of the picture. A system could be created by one person, operated by another, and exploited by a third.
To understand what had happened, Elena needed to trace the entire process.
Who decided that certain complaints posed a risk?
Who determined which files should be moved?
Who approved the transfers?
Who contacted witnesses?
Who benefited when the official record became incomplete?
And who had the authority to stop the process but chose not to?
She was still considering those questions when Priya called again.
“We found the transfer sheet.”
Elena stopped writing.
“Is it the one Samuel Greene described?”
“It appears to be. The document contains his initials and a date matching the period he mentioned.”
“Does it list the eleven red-coded individuals?”
“Yes.”
“Have we identified them?”
“Some. Several names match complaints already under review. Others are new.”
“Any indication of what the colors mean?”
“Not yet. But the file contains a separate legend.”
Elena held her breath.
“What does it say?”
Priya read from her notes.
“Green means standard administrative handling. Yellow means elevated reputational sensitivity. Red means executive notification before further contact.”
Elena closed her eyes.
The language was bureaucratic, but the meaning was clear enough to justify deeper scrutiny.
The system had not merely tracked the status of complaints.
It had classified people according to the potential risk they represented to the institution.
“What happened to the red-coded complaints?” Elena asked.
“That is what we are trying to establish.”
“Compare every name against the official case history. Look for unexplained closures, transfers, missing recordings, and contact with employers or other third parties.”
“We will.”
“And find Thomas Bell. He may be the only person who can explain how the classification system was used in practice.”
Priya agreed.
After the call, Elena returned to the whiteboard.
Beneath the three questions, she wrote a fourth.
WHO IS R. K.?
She stepped back and studied the words.
For months, the investigation had been driven by fragments: missing documents, vague warnings, unexplained transfers, and a witness who had vanished.
Now the fragments were beginning to form a picture.
But a picture was not the same as proof.
Elena knew that the next phase would be the most dangerous. The people whose decisions had shaped Project Northstar would understand what the archive could reveal. They might attempt to discredit the evidence, challenge its authenticity, or portray the investigation as an attack on the entire department.
They might also try to silence the witnesses.
Elena took a fresh sheet of paper and began drafting a list of immediate priorities.
Preserve the archive.
Authenticate the documents.
Protect the witnesses.
Trace the money.
Identify the people who created and used the system.
And never allow the investigation to become a story about one woman confronting one angry officer.
Because if Project Northstar had done what the documents appeared to suggest, then the sidewalk encounter was only one visible moment in a much longer history.
A history in which people had been classified, monitored, and dismissed.
A history in which the official record had not always reflected what actually happened.
A history in which those who complained had sometimes been treated as threats.
Elena finished the list and placed it beside the photograph of Lakefront Document Services.
She did not know how long it would take to uncover the entire truth.
She did not know whether Thomas Bell would come forward, whether Officer Harrison would cooperate, or whether Robert Vance would attempt to obstruct the investigation.
But she knew the evidence was finally moving into the open.
And once the truth had a record, a timestamp, and a chain of custody, it would become much harder for anyone to pretend that it had never existed.