8. Cali #2
No imagined version included the small, specific, beautiful details.
The way his mouth keeps moving for half a second after his eyes have locked on, like a dying balloon running out of air.
The hand stuck in mid-air, halfway toward Glover.
The pale tint of his face.
And then, like fuel on the fire that's about to burn him to a crisp, the proof of where he's been for the last hour. A faint hickey under his jawline. A lipstick stain on his collar.
And under the bank's recycled air, faint but unmistakable as it drifts toward me, the floral, warm, powdery perfume that has lived rent-free in my nightmares since that fateful afternoon in the school pick-up line.
Sweet Spun Vanilla Betrayal.
He's still wearing her.
"… Cali?" he says.
"Mr. Bradshaw," I say warmly. I extend my hand across the table like he's a stranger. "So glad you could join us."
Glover, who knows nothing, claps his hands together. "Great, you two know each other? Even better. Todd's worked with my firm before. Solid guy. Todd, sit, you're going to want to see this. Cali's a rock star in the making."
Todd clearly does not want to sit. His entire body has the trapped, swiveling quality of a rat cornered in a dark alley.
But all of his colleagues and one of his biggest clients are all looking at him expectantly, and a man as prideful as Todd cannot do the one sane thing, which would be to turn around and walk out.
He's stuck.
So he sits.
The fresh hickey on his throat bobs as he swallows. "I—didn't realize you were—presenting," he manages.
"Surprise," I say sweetly. "Don't worry. I think you're going to absolutely love it. It's got a ton of application to your area of focus."
He tries to say something—I think it's "Wonderful"—but the way it comes out through his dry, sticky throat is more like the last cough of a man dangling from the gallows.
"Let's get started, shall we?" Emmett declares, clapping his hands. "I've gotta be on a helicopter by one."
"Of course!" I reply. "Let's get right into it."
The junior banker who's been helping me dims the lights, and all the senior bankers settle back in their seats to watch and listen.
The dashboard lights up on the wall screens behind me in neat blocks of sober blue and gray.
Clicker in hand, I turn to face my audience.
"Thank you all for taking the time to meet with me.
What I'm going to show you today is a predictive-analytics engine that ingests disconnected, multi-source, timestamped data and finds the story hidden inside it.
Most fraud isn't one big lie. It's a hundred small ones that only become visible when you correlate them. "
The tech guys lean in. Glover folds his arms, intrigued.
Todd sits very, very still.
"The classic use case is vendor fraud." I walk them through the same story I prepared for Phoenix, the one that initially caught Glover's eye.
It's a sanitized supply-chain demo, involving phantom shipments and an invoice that doesn't match a delivery that doesn't match a GPS log.
As I explain the ins and outs of the technology, the engine catches it in seconds, lighting up the anomalies in bright, unmissable red.
"On their own, each data point is innocent," I say in a loud, clear voice.
"A late truck. A rounding error. Could happen to anyone, right?
" I click through to the next slide. "But the human eye can't hold this many data streams at once.
The engine can. It correlates the streams, surfaces the pattern, and outputs one clean, undeniable picture. "
"Now," I say, clicking forward again, "of course, this is merely an illustrative example.
But you all know that lots of pretty tools don't survive contact with reality.
So let me show you what happened when we ran a live, real-world dataset through it.
Something messier than vendor invoices. Something painfully human. "
With the push of a button, the supply-chain slide dissolves. In its place, a fresh timeline begins to draw itself across the screens from left to right, eighteen months wide.
I keep my voice cool and instructive. Emotion won't help me here. These men want facts.
Lucky for them, I have plenty.
Eight, to be specific.
"The subject here believed his activity was untraceable.
He believed that because each individual data point was, on its own, perfectly mundane.
A school pickup line sign-out log." Click.
"A song played on a shared account at 2:51 in the afternoon.
" Click. "A smart-scale reading from a bathroom that should rightfully have been empty. "
The room is attentive. Somebody's chair stops squeaking. I have a very good guess whose chair that might be.
"Each one looks innocent," I say. "Easily explained away. But, as I said, the human eye can't hold them all at once."
Click.
"… The engine can."
The streams snap toward each other across the timeline. I turn to face the table, and I let my gaze land, just for a heartbeat, on Todd.
"Is this built off vendor data?" another MD asks.