Contractors’ pay applications are getting a new reader. Fort Hill LogiX, spun out of Greenville, S.C., construction audit firm Fort Hill Associates, launched an AI-assisted engine on Oct. 15 that reviews payment applications for capital project owners and flags line items that don’t match the executed contract.
The pitch is aimed squarely at owners, not builders. Fort Hill Associates has audited more than 1,600 capital projects since 2006, mostly for universities, health systems and similar institutions, and it’s the new tool’s first customer.
How the AI construction audit tool works
The design is more conservative than most construction AI launches, and that’s a point in its favor. AI reads the project documents. Fixed rules, not the model, test each line item against the contract. Each exception comes back with the page where it was found and the supporting document attached, and a Certified Construction Auditor reviews every exception and signs the report before the owner sees it. The owner makes the final call.
The company says the process follows the National Association of Construction Auditors’ Construction Audit Professional Standards. The examples it gives are the familiar ones: billed clauses that don’t match the contract, labor rates that don’t reconcile and change orders that don’t line up with the agreement. All four founders hold the CCA credential, according to the release.
Founder Ruhtab Sahota said the team didn’t set out to build software. Co-founder Curt Plyler, a CFA charterholder, framed the product around owner risk on capital tied up in a project.
What isn’t in the launch
Plenty. LogiX hasn’t named an outside customer, a price or any accuracy figures, and it hasn’t said how many pay apps it has run through the engine. Its only user so far is the audit firm that built it. That isn’t unusual for a first release, but it means the claims about speed and coverage are untested by anyone without a stake.
Accuracy will be the real question. Exchange covered Togal’s takeoff benchmark last week, where the best general AI model scored 34% on construction drawings. Splitting reading from rule-checking, with a human signing off, is a reasonable hedge against that kind of error rate.
Why GCs should care about pay application review
Owner-side audit has historically been sampled and slow, often done at closeout when the money’s already out the door. A tool that checks every line on every draw changes the timing. Disputes over labor burden, markup on change orders and general conditions would surface monthly instead of at the end.
For general contractors and subs, the practical response is housekeeping: rate sheets that tie to the contract, change order pricing that follows the agreed markups and backup attached to every line. Plenty of firms already work that way. The ones that don’t will start hearing about it every 30 days, with page numbers attached.