Priced from the jobs you built.
Unit rates from your own completed work — 61 closed-out jobs on a bid like this — escalated for time, adjusted for the productivity your crews actually achieved, and re-based for region. Not a national index, and not last year's spreadsheet.
What it does
Every job your firm has finished is a priced experiment: what the scope cost, which crew installed it, how long the program ran, what the region was doing to labour that year. Cost AI prices the bill of quantities off that record line by line, with the source jobs listed beside every rate.
Rates are adjusted, not copied. A feeder rate from a 14-month tower in another market is escalated for time, adjusted for the productivity that crew actually achieved, and re-based on region before it is offered. Checked against awarded cost, the estimate lands within ±3.4%. The estimator changes what is different about this job instead of starting from a blank sheet.
How it works
- 01
Read your finished jobs
Awarded values, final costs, installed quantities, crews and program lengths from every job you have closed out — the same record production tracking and the cost ledger already keep.
- 02
Match the scope
Bill-of-quantities lines are matched to work you have built before by system, trade and installation method rather than by keyword. Every candidate rate names the jobs it was drawn from.
- 03
Adjust, then offer
Each rate is escalated for time, adjusted for achieved crew productivity and re-based for region, then offered with its sources. Overrides are kept with the estimator's reason and train the next bid.
Spec
- SOURCE JOBS
- 61 CLOSED OUT
- ACCURACY
- ±3.4 % VS AWARDED
- RATE SOURCE
- OWN HISTORY · NOT INDEX
- OVERRIDES
- KEPT WITH REASON
±3.4 %
Estimate against final awarded cost