The control room door opened on my biometric, and I stepped aside to let her in. She stopped just past the threshold. I waited for the reaction. The wide eyes, the slow scan, the questions about what everything did and how much it had cost and why a man who supposedly consulted on tech infrastructure needed a setup that rivaled a government facility.
Instead, she walked to the secondary desk where I’d set up her workstation and ran her hand along the edge of the monitor. “Dual screen?”
“I figured you’d want the model on one and reference data on the other.”
“That’s exactly how I work.” She sat down. Adjusted the chair height. Moved the keyboard half an inch to the left. Precise, automatic, the way I positioned my own equipment every morning. “This is how you spend your days.”
It wasn’t a question, but I answered it anyway. “Most of them.”
“I built something like this at the Bureau. Three monitors, networked storage, a workspace nobody else used because the cubicle was too far from the break room.” She looked around once more, and what I saw in her face wasn’t admiration. It was recognition. The expression of a person who understood what it meant to construct a world that fit your brain. “Yours is better.”
“Mine cost more.”
She almost smiled. I almost let myself enjoy it.
We started with her framework. She loaded the rebuilt model onto the system, and I pulled my intercept data into a parallel feed.
“This cluster here.” She pointed at a concentration of courier movements south of Kalispell. “My model predicted activity in this window, but the frequency is higher than I projected. Why?”
“Kindt shifted three additional couriers into the western corridor about six weeks ago. Summer traffic on the highways gives them more cover.”
“Seasonal adjustment.” She was already typing, updating her variables. “That would also explain the deviation in the northern routes. I assumed it was a personnel change, but if it’s traffic-based cover, the whole model needs a seasonal weighting layer.”
“How long to integrate that?”
“Give me twenty minutes.”
It took her twelve. She rebuilt an entire variable layer in her framework while I verified against my intercepts, and the revised model snapped the corridor data into alignment like a lens coming into focus.
“That’s a ninety-one percent match now,” she said.
“What’s the nine percent?”
“Noise, probably. Or a route I haven’t mapped yet.” She was already scanning the gaps.
She pulled up a secondary layer on her screen. “There’s also an external variable I’ve never been able to resolve. Kindt’s pipeline has been disrupted multiple times over the past year by something that isn’t law enforcement and isn’t a competitor. I’ve been working around it, but with your intercept data I might be able to finally identify the source.”
My chest went tight. I knew exactly what that external variable was. Me.
“How much is it affecting your projections?”
“A few percentage points. It’s not critical. Just irritating.”
“Then park it. The route gaps are more important right now.”
She gave me a look—the one that said she’d heard me prioritize and was deciding whether to argue. Then she nodded and went back to scanning the gaps.
“Here. This southeast branch. My model predicts courier activity, but your intercepts show nothing. Either they’re using a communication channel you haven’t found, or my projection is wrong.”
“Or they’re running that route dark. No comms. That’s what Kindt does with the youngest kids. The ones worth the most to buyers. No digital footprint, no scheduling chatter, just a van on a back road that nobody’s tracking.”
She went very still. “How young are we talking?”
“The chatter I’ve intercepted references children as young as four.”
She turned back to the screen. Her jaw was tight, but her voice was steady when she replied. “Then we need to pinpoint that route.”
“Agreed.”
