2. Vincent
VINCENT
Ifind the anomaly myself, which is the first problem.
The variance is small. That is the first thing that tells me it is deliberate.
I pull the full inventory thread going back sixteen months.
The pattern clarifies. Small draws from three different nodes, all below the alert threshold, all targeting drugs from discontinued trial lines.
Not high-street pharmaceuticals. Not the compounds with resale value on secondary markets.
Someone is taking the drugs nobody is looking for, because those are the drugs nobody is watching.
KD-9. A suppressed vascular treatment from a trial we closed fourteen months ago. The math is not difficult. The discontinuation preceded the first theft by approximately three weeks.
I set that down in my mind carefully, the way you set down something fragile. I do not draw the line yet but I note the proximity.
Graham answers on the second ring, which tells me he was already awake. He always is. It is one of the qualities I find most useful in him.
"Internal review," I say, without preamble. "Supply chain. I want it run from your office, no external auditors, no compliance flags sent upward until I have seen the full picture."
A pause. Not hesitation, Graham does not hesitate. He is processing scope. "Timeline?"
"Two weeks to preliminary findings. I will send you the distribution thread tonight."
"Who else knows?"
"You do now."
He says he will have a framework by morning. He does not ask why I am running this through legal rather than operations, and I do not explain it. That is the other quality that makes him indispensable.
I stay at my desk for another hour after the call, not because there is more to review. I have extracted everything the data can tell me tonight. But because I am thinking about the mind behind it.
Theft at this scale and precision is not driven by need.
It is not driven by ideology, not at its core.
It is driven by the confidence of someone who has studied a system long enough to know where it is blind.
I have spent my career building those blind spots deliberately, to protect against exactly this kind of incursion, and someone found them anyway.
I am not angry. I extend it the same quality of focus I reserve for problems that have earned it.
The ER report arrives in my inbox at 11:47 p.m., forwarded from Owen's assistant. Owen is my Senior VP of Regulatory Affairs and, as of approximately three hours ago, a patient at St. Aurelius Medical Center following a high-speed collision on the Hartwell overpass.
The assistant's message is four sentences of appropriate alarm followed by a note that the attending physician described his prognosis as favorable and requested that all further clinical inquiries go through the hospital's patient services line rather than directly to her department.
The phrasing is careful enough that I read it twice.
I open the attached incident summary. The attending physician on record is a Dr. Clara Whitlock, emergency medicine, twelve years at St. Aurelius.
The summary notes, in flat administrative language, that Dr. Whitlock removed two members of Owen's private security detail from the trauma bay upon arrival, citing sterile protocol requirements.
She delivered the preliminary prognosis to Owen's assistant by phone herself, without being asked, in under ninety seconds.
Stable, surgical intervention likely, recovery timeline pending imaging.
She then apparently redirected all subsequent calls to patient services.
I read it a second time.
The detail that stays with me is not the removal of the security team.
That was protocol-correct. They had no business being in a sterile bay, and whoever sent them in did not think past the reflex of proximity.
What stays with me is the unprompted prognosis call.
She did not wait for someone to ask. She identified what information was needed, transmitted it with clinical efficiency, and closed the channel.
She handled the problem of the anxious assistant the way she clearly treated everything else in that bay, as a variable to be resolved so she could get back to work.
I set the report aside. In the same mental space where I keep the distribution anomaly — the theft pattern, the discontinued trial, the proximity of dates — I place nothing about Dr. Clara Whitlock. She goes somewhere else. A different drawer entirely, closed with more care than the others.
It is not a decision I examine. I simply make it, the way I decide. Cleanly, without deliberation.
I pull up Graham's email thread and attach the distribution data. The investigation will take the shape I have specified. Quiet, contained, thorough. Whoever is behind the theft understands my system well enough to steal from it invisibly. That narrows the field considerably.
Medical training, likely. Clinical knowledge of discontinued trial compounds, certainly. Access to distribution routing data, or someone who has cultivated a source with that access. The profile will sharpen once Graham's team starts pulling the thread.
What I keep returning to is the selection.
KD-9 is not a drug with street value. It has no secondary market, no recreational application, no obvious resale pathway.
Whoever is taking it knows exactly what it does and has a reason to want it that has nothing to do with profit.
That reframes the entire exercise. This is not theft in the conventional sense.
This is redirection, someone correcting a supply chain outcome they have decided is wrong.
That kind of conviction is not common. Neither is the operational discipline required to act on it over fourteen months without a single detectable error.
I have built this company on the premise that systems are more reliable than people.
Whoever is behind this appears to operate on the same premise, applied in the opposite direction.
I close the supply chain file.
I open the ER incident report one more time. Read it straight through. The attending physician is described, in the assistant's phrasing, as having been "quite direct." I imagine this is a diplomatic rendering. I find, somewhat against my expectations, that I have no interest in that.
What the report contains is this: Dr. Whitlock removed Owen's security detail before she had his name.
She called the update through to his assistant unprompted, in under ninety seconds, then redirected the line.
She did not wait for permission to act and did not seek acknowledgment after.
The report is six paragraphs of clinical notation and one line from the assistant that reads, verbatim: she didn't seem to care who he was.
That line is the one I read a third time.
Most people, when confronted with the resources and proximity that surround Owen's position, adjust. Not dramatically, they rarely realize they are doing it.
But something shifts in how they speak, how they hold themselves, how much weight they assign to the outcome.
It is an almost universal response to power, and I have spent enough time in rooms where it operates to recognize its absence immediately.
She apparently did not adjust at all. She assessed the patient, removed the obstruction, and got back to work.
I close the file. I do not examine why I have now read it four times.
Outside, the logistics of a long night settle into the deep quiet of a building that has exhaled and gone still. I make tea — single origin, same preparation it has always been — and sit with the two things I have filed, in their separate drawers, not touching.
The thief, who mapped every blind spot I built.
The physician, who did not blink.
Two different problems, I tell myself. Two different rooms.
I drink the tea. The drawers stay closed.