2. Tyler #2
It's good. Not adequate. Not passable. Good in the way that makes the back of my neck hot because I can't find the flaw I need to justify killing it.
Her redundancy protocols stay intact, nested inside a tiered processing model that offloads the heaviest computations to a distributed node system her team built from scratch.
Elegant. Efficient. And completely incompatible with a Q3 launch.
I call Dawson.
"You looked at her proposal?"
"Twice. It works, Tyler. The stress tests came back clean. No degradation up to ninety-thousand concurrent queries."
"And the timeline?"
Silence. The kind that answers the question before words do.
"Fourteen weeks minimum to implement her framework alongside ours. Could stretch to sixteen if we hit snags during the data migration."
Fourteen weeks puts us into Q4. Past the product launch window.
Past the earnings call where I promised three hundred institutional investors a live demo of our integrated AI platform.
Past the point where Meridian Capital, our largest competitor, rolls out their inferior but faster product and captures market share we'll spend eighteen months clawing back.
"Dawson. Can you compress it?"
"Not without pulling resources from two other active projects. And even then, twelve weeks. Maybe."
Maybe. The most expensive word in the English language.
I end the call. Open Shayla's proposal again.
Scroll to page nine, where she's outlined the custom middleware her team needs to build, the bridge between her architecture and ours.
It requires a dedicated development sprint.
Four engineers. Eight weeks. And a budget line item of $340,000 for cloud computing resources to run continuous integration testing during the build.
$340,000. Rounding error on a company this size. Enough to fund her middleware. Enough to keep her timeline alive.
Enough to guarantee I miss Q3.
I close the document. Open the departmental budget dashboard. Bring up the allocation for Project Nightingale, the internal codename for her AI integration. The middleware sprint sits under a sub-line labeled "Custom Bridge Development." I highlight it.
My cursor hovers.
This is the part they never write about in the profiles.
Not the deal-making. Not the vision. The math.
The clean, bloodless arithmetic of choosing which limb to cut so the body survives.
Three hundred investors don't care about redundancy protocols.
They care about returns. The board doesn't care about neural network safety nets.
They care about the stock price between now and January.
And Shayla Barnes, for all her brilliance, for all the ways her mind works in dimensions I haven't seen since I was young enough to still be impressed by intelligence, does not understand that her sixteen-week timeline is a luxury this company cannot afford.
I rewrite the budget. Strip the $340,000 cloud allocation entirely.
Reduce the engineering headcount from four dedicated resources to one shared resource, a junior developer from Dawson's team who'll split time between three projects.
The middleware sprint, her bridge, becomes functionally impossible.
Without it, her alternative architecture can't integrate on her terms. She'll have to fall back to Dawson's original framework: strip the redundancy protocols, run parallel systems, and hit the ten-week window I set on day one.
Checkmate dressed as a spreadsheet.
I draft the email. CC Dawson, Grace, the project management office, and the VP of Finance.
Subject line: "Project Nightingale: Revised Q3 Budget Allocation.
" The body is three paragraphs of corporate language so polished it could pass for diplomacy.
"After careful review of resource constraints and strategic priorities...
" "Regrettably, the current fiscal quarter does not support...
" "We remain committed to the successful integration of all acquired technologies within the established timeline... "
Every sentence is a locked door. No ambiguity. No room for appeal. The decision flows from my authority as majority stakeholder and CEO, which means the only person who can reverse it is me.
I read it one final time. The cursor blinks at the end of my signature line.
Through the glass wall, I can see her office. She's not there yet. 6:53 AM. The lamp is off. The mustard throw hangs undisturbed over her chair. That ridiculous plant catches the early gray light from the window, its waxy leaves still and green and obscenely alive.
In seven minutes she'll step off that elevator. She'll flip on the lamp. She'll set down whatever homemade meal she's packed for the day. She'll open her email.
And she will come for me.
The thought produces a sensation I choose not to examine. A wire pulled taut somewhere beneath my ribs, vibrating at a frequency I don't have a name for. My mouth is dry.
I click send.
The email disappears from my drafts. Delivered. Read receipts on. I lean back in my chair, lace my fingers across my stomach, and face the glass wall.
6:54 AM. The elevator hasn't chimed yet.
I wait.