My brain paused, “How early?”
“Yes,” she said, “What’s the average prediction window?”
My eyes flicked to the graph. The answer was there in the numbers but the sentence tangled in my head.
“Approximately… uh… Approximately two hours… before… before the…”
The word slipped away.
Transaction.
Why couldn’t I grab it?
“Before transaction authorization,” Raphael said calmly.
“Yes,” I said quickly.
“Before transaction authorization.”
I clicked to the next slide too fast and another man leaned forward, “What behavioral indicators trigger the initial score increase?”
“Spending… location… and… device—”
The sentence collapsed halfway through, my eyes darted to the screen. The letters blurred together.
Raphael spoke again, “Primarily location deviation and unusual spending velocity,” he said.
The client nodded, “That makes sense.”
I tried to continue, “Yes, the system compares… compares…”
My fingers hovered over the keyboard like that would somehow help.
Another client spoke, “How does the model avoid false positives with frequent travelers?”
My mind went blank. Frequent travelers. I knew this. I had written the algorithm but the words wouldn’t line up.
“It adjusts for… for… uh…”
Raphael stepped in again.
“The system builds a dynamic behavioral baseline,” he explained smoothly. “So if a customer regularly travels, the algorithm adapts to that pattern.”
The clients nodded again.
I could feel my chest tightening more with every second. I tried to continue the presentation but now the sentences tangled faster.
Words flipped.
My mouth started forming half thoughts that didn’t land properly.
“The system… detects… behavior pattern… difference… before…”
I heard myself say it and knew it sounded wrong.
The man at the end of the table frowned slightly, “I’m sorry,” he said politely, “Could you clarify that?”

VERIFYCAPTCHA_LABEL
Comments
The readers' comments on the novel: My Stepbrother's Dirty Little Secret