Page 26 of Thorne
I drop my head into my hands, and the weight of the silence presses in.
Then I let the numbers do what they always do.
Not counting. Never counting.
The rules begin moving through my head in the quiet, mechanical sequence I learned as a child, nearly the same age Lily is now: double the digit, add the neighbor, subtract from nine, carry the remainder forward. Each step triggers the next before the previous one has fully settled, the structure assembling itself in layers behind my eyes.
The Trachtenberg patterns don't hesitate. They don't reconsider. Once the sequence starts, it runs to completion exactly as designed.
Clean.
Predictable.
Unaffected by anything happening in the room.
My lips move as the operations stack in my head, the words barely audible.
Double … Neighbor … Subtract …
The rhythm steadies my breathing. Each rule locks into the next like gears engaging, the calculation unfolding with the quiet certainty of a system that has never once cared about fear, or anger, or the weight of a child waving from across a yard.
Numbers are honest. They do exactly what they're told. They stay where I put them.
And then, because my mind does what it does when given enough quiet, the comfort of math tips over into the actual problem.
They need the patient distribution manifests for the ML-273 trial. Every adult and child who received the compound. Here's what they don't understand yet, because I haven't told them, and they haven't asked the right question: there is no list to retrieve.
Phoenix designed the distribution architecture with one absolute security constraint. The master patient directory was never committed to any system it didn't control. No external backup. No accessible archive. The distribution data lives inside Phoenix's encrypted network and inside one other location.
My head.
Not the names. Phoenix kept the names. The specific coordinators, the specific clinics, the specific children. Phoenix was careful about that. A human architect who holds the full directory can defect with it.
I created the design, but the directory was always Phoenix's.
What I hold is something more foundational and, for their purposes, more immediately useful than a list of names I don't have access to.
I built the model that determined what a qualifying regional coordinator required. The precise institutional fingerprint each distribution hub had to match.
Geographic coverage radius.
Patient volume.
Post-remission trial enrollment capacity. The financial intake structure required to absorb Meridian's funding without triggering reporting thresholds.
I built those parameters with the same meticulous detail I bring to every financial architecture, because Phoenix demanded precision, and because I'm exceptional at my job. Because I told myself for a long time that being good at the work meant the work was good.
I was wrong about the last part.
But the parameters are intact. Every one of them.
The team already has one confirmed institution. I gathered that from the briefing before they locked the door. A Northeast anchor. A children's hospital. CHOP meets the requirements of the Northeast tier. That match is not a coincidence. It'smy model functioning as designed. One confirmed institution means the profile holds. It means I can describe the remaining sites with sufficient precision for them to be found.
I begin the only way I know how: from the structure outward.
Geographic dispersal engineered to prevent clustering that would trigger CDC epidemiological flags. The Northeast anchor carries the highest volume: a major research hospital with a dedicated immune-recovery ward and a post-remission cohort capacity of at least fifty patients.
The Southeast anchor requires proximity to high-density transit hubs, utilizing a decentralized financial intake structure to absorb Meridian's funding without tripping IRS reporting thresholds. The Midwest tier needs a specific ratio of rural-to-urban trial access to mask the compound's true distribution curve …
