A rep's request and an equipment schedule, read line by line into a quote sheet for your applications engineer.

A request comes in from the rep agency as an email with attachments: the schedule page from the mechanical drawings, the spec section for that equipment, sometimes an addendum and a list of approved alternates. Your applications engineer pulls each tag's airflow and static, coil and fan data and electrical data, notes the construction options and tests the section adds, and only then opens the selection program.

In the test, AI reads the same copies and drafts the quote sheet: unit tag, airflow and static as scheduled, coil and fan fields, voltage, phase and ampacity fields, the construction options the spec calls for, the testing or certification it requires, and the lines a person has to select or resolve, each with the page or clause it came from. It runs no selection, prices nothing and sends nothing to a rep.

The tag is the key, and owners don't write the line the same way. The University of Georgia's standards, dated July 1, 2026, make every schedule give each piece of equipment a unique identifier, rule out a single type standing in for several units and say a building can't hold two AHU-1s. They also want the basis of design by make and model with all options, once the designer has verified the model is still in production. Old Dominion University goes the other way: it discourages designers from printing a manufacturer's cooling, heating and efficiency data on the schedule, so makers compete on calculated values and code minimums, and it allows manufacturer values only for custom and specialty equipment. One request names a competitor's model with every option. The next gives only the duty. The draft carries each line as written and says which kind it read.

Static pressure depends on who added what. Old Dominion has every scheduled fan with filters state its dirty filter pressure drop allowance. Georgia bases fan static on filters half loaded, plus system effect, internal cabinet losses, external duct losses and the drops through coils, dampers and humidifiers. The University of Houston's section for custom air handlers then moves the target: the fan has to run stably at half an inch above the scheduled static, and the motor can't overload a half inch below or above it. Old Dominion even fixes elevation, letting design calculations use sea level or 17 feet. So the draft records external or total static as the sheet states it, the filter condition, any margin the spec adds and any elevation it gives, and leaves a question where the request is silent.

Coil and fan lines go well past the schedule. Old Dominion wants each designed coil's entering and leaving air and water conditions, air and water pressure drop and tube velocity on the schedule, with fouling factors alongside. AHRI's coil program certifies average total capacity, air pressure drop and fluid pressure drop, so the rest is your own construction data. Houston's master section offers bracketed choices of 8 or 12 fins per inch, tube walls of 0.020, 0.024 or 0.035 inch and 304 or 316L stainless casings, and tests coils by air pressure under water at one and a half times the pressure on the drawings, never below 250 psig or above 500. A bracket left unedited is a question, not a value. On the air side, Houston's fan array paragraphs call for a backdraft damper on each fan and one complete spare fan and motor assembly with every unit, while Georgia specifies direct drive fans with variable speed drives and puts belts behind a written variance.

Certification lines look alike and aren't. AMCA warns that a product tested to an AMCA standard isn't necessarily certified, that certified makers aren't required to show the seal, and that a requirement to bear the seal can be met with literature showing it in a submittal. Houston wants fans that bear the seal for air and sound. A 2023 elementary school section from Massillon City Schools in Ohio wants air handlers cETLus listed to UL 1995 and CAN/CSA C22.2 No. 236, and certified under AHRI's Standard 430 program with a note excepting fan arrays. Its filter line says UL 900 listed, Class I or Class II, where Houston's says Class 1. The same Massillon section limits casing leakage twice, once per square foot of casing and once as a share of supply air, at different test pressures. On a California hospital, certain equipment needs a Certificate of Compliance for special seismic certification, which the state's voluntary OSP program reviews and preapproves. The draft lists each requirement in the owner's words, takes no position on whether a unit meets it, and marks any two that conflict.

The expensive sentences sit away from the schedule, and your reviewer watches for them. Houston has the owner's representatives witness factory vibration, acoustical and unit run tests, with the contractor paying air fare and lodging for two of them. Old Dominion asks for factory and field leak testing on air handlers above roughly 10,000 cfm, with witness testing coordinated through its facilities office. Massillon says that if a submitted unit differs from the design basis in size, capacity, connections, electrical needs, weight or pad requirements, the manufacturer bears the cost of the changes and of coordinating the other trades. Your reviewer, usually the lead applications engineer, checks that every tag came across, no scheduled number changed, no value was invented for a blank or a bracket, every test and listing the section names landed on a line, and every clause like these reached the list a person has to resolve.

A yes from your reviewer moves the same reading onto new rep requests for about 30 days.

Every quote still leaves from your own system after your applications engineer has selected, checked and approved it, and the scoring table keeps running on live requests. Anything further is a separate decision with its own price, with monthly support after that.