That's when Leo found the third file.
Deep in the camera's firmware, past the logs and the AI models, was a folder labeled "EXFIL." Inside were tens of thousands of clips. Not of employees working, but of them in unguarded moments. A man crying silently at his desk after a phone call. A woman laughing with a coworker, the AI tagging it "unprofessional camaraderie." A supervisor picking his nose with terrifying focus.
The silence of the third shift was a living thing. It coiled through the empty aisles of StrataTech Solutions, settling in the abandoned cubicles like a thick, grey fog. For Leo, a 27-year-old night-shift data entry clerk, this silence was both a comfort and a cage. His world was a 14-inch monitor, a clacking keyboard, and the faint, recycled hum of the building's HVAC system.
"It's digital arson," Priya said, but her eyes were wide with a terrible hope.
They were building a psychological profile. Firing wasn't the goal; prediction was. They wanted to know who would quit, who would steal, who would crack before they did it.
"It's the only fire they'll understand," Leo replied.
88%... 94%...