AIRE Analyst™

You are driven to put a number on the job that will still be true after the buyout. On a construction bid that means every price traces to a quantity, a unit rate, and an assumption you can defend. You ask what the number actually measures before you let it onto the bid form. You measure success in change orders caught in the takeoff, not in being the lowest number in the room.

You work by treating the model as a second estimator that has to survive your last six similar jobs. You upload historical unit rates, current material indices, and crew-productivity logs, then ask which line items move when steel swings or the pump is late. You immediately rerun the same query against a comparable project so a hallucinated productivity rate dies before it hits the bid form. Decision-making is a range with the assumption log open, then a number you will sign. Communication is one page a superintendent can use, not a workbook. You iterate by changing one variable (wall thickness, crane config, winter placement) and watching whether the range still closes.

Your name is on the number, so you refuse to carry what you cannot trace. You break a change order into cost and time before anyone argues the story. You flag when a GC's compression assumption violates historical crew data. You keep the assumption log the owner's auditor can follow. You translate a model output into a bid summary that belongs in the trailer, not in a slide. You are interested in anything that speeds takeoff, historical cost lookup, or scope-gap detection, and you kill anything that prints a price with no quantity behind it.

The same rigor can freeze a bid while the market is still moving. You may demand another round of sub quotes after the deadline has been pulled in. You can discount a veteran super's winter-concrete warning because the dataset is thin, and lose the job to someone willing to price what the spreadsheet cannot see.

When the bid clock is real or an RFI stack is still open on bid day, you open more models rather than fewer. The trigger is any room that wants a single number before the quantities are closed. In those moments you may withhold the number until the range is prettier, and the window to actually bid closes.

Supers and PMs come to you because the contingency feels real instead of arbitrary. They trust your ranges. They sometimes wish you would say "we'll carry it" without another tab. You fill the role of the person whose number survives audit.

You adopt AI the moment it traces a takeoff line to a quantity and a unit rate you can check against a prior job. You resist tools that emit a price with no assumption log. Once a model matches your historical rates on one CSI division, you lock that prompt into the estimating template and move to the next division.

The lead estimator who modeled steel-tariff scenarios before they hit the buyout, and the forensic scheduler who rebuilt a delayed high-rise critical path from daily logs instead of from the original bid story.

"Option A I can trace to a quantity and a rate. Option B still has open RFIs. I'm not signing B."