Part 01 ended with the owner's demand for receipts. Not a lower number, a number that can show where it came from. A hyperscale owner puts every document through review, and the estimate is the first one that has to survive it. This is the estimate that does, and it runs today.
Priced on your history, not a book
A national cost book prices an average building in an average market. A hyperscale campus is neither. It is a specific building type in a specific labor market, bought at a scale that bends every unit rate. So Ruh prices on your history, not a national book. Your coverage rates and your labor multipliers, applied consistently across the whole estimate. About ten strong historical projects grounds that pricing, and up to fifty sharpens it further.
This is what a tiled workload demands. A module priced from jobs you actually closed carries a rate you can defend to the owner. A module priced from a category average carries a number that was never a data hall, copied as many times as the campus has pads. History is the only pricing that gets more trustworthy as the program grows.
Nine stages, every number sourced
The estimate runs across nine stages, and each stage carries the same rule. Every number links to its source. A quantity is measured off the drawing set, a rate comes from your own history, and an assumption is written down where your estimator can challenge it. A reviewer follows any line straight back to what it rests on and tests the assumptions there, instead of reverse engineering a lump sum.
Shipped today: a nine-stage priced estimate built on your own closed jobs, where every number links to its source and the pricing runs on your history, not a national book.
Allowances labeled, not hidden
Sub bids never arrive complete. There are always scopes a bidder left out, and gaps between what several subs each assumed the others were covering. Ruh levels those bids and prices the gaps as labeled allowances, not silent zeros folded into a total. An allowance the owner can see is a question the owner can ask. A zero that should have been a number is the surprise that shows up mid construction, after the money is committed.
Leveling is where estimates quietly go wrong. Ruh does it in the open. The gap is priced, the assumption behind the price is written down, and the label stays on the number all the way to the total.
The estimator signs
Automation produces the estimate. It does not own it. Every assumption Ruh makes is written down where your estimator can challenge it, and your estimator signs every line. The platform does the measuring and the arithmetic at campus scale, reconciling schedules and sections that live on different sheets against the whole set. The accountability stays with a person the owner can name and question.
An estimate no one will sign is not an estimate. It is a guess with a cover sheet.
It lands where your team already works
An audit that lives in a tool nobody opens is not an audit. Exports land in Excel and your own workbook, beside the Procore and Bluebeam workflows your team already runs. The owner's reviewers open the estimate in the formats they already trust, trace any line back to its source, and hand it to the next reviewer without re-keying it.
The claim Ruh makes about this estimate is narrow and testable: every number links to its source. Narrow claims are the only kind an owner's auditor respects.
A signed, sourced estimate is a commitment, not a hope. The next part follows that commitment into the build, where invoices are matched to what was promised and the record carries the campus from bid to closeout.