Estimators put AI to work from the site visit to the final review

Chris Brasington spent 10 years in construction and worked in landscaping for eight more, helping companies automate. The biggest loss of time in estimating, he told an Ignite session, is finding, organizing, and re-entering information. He sees estimators “start digging through data, and they dig, and they dig, and eventually they may find it, they may not — then they start estimating.”

The Ignite 2026 session, “Real-World AI: How Estimators Are Cutting Bid Time in Half,” paired Brasington, founder and technology strategist at Global Solutions, with Ramel Sanchez, founder and principal of Miri Consulting. Joshua Martin, group product manager at Aspire Software, moderated. From there, the conversation followed an estimate from the first site visit to the review afterward.

AI is changing the way estimators operate

The site visit came first. An irrigation company Brasington works with takes video of the walk-through with the client on a phone. An AI transcriber turns the video into text, and the company’s AI searches previous estimates and returns the five closest matches. The estimator starts from those. Martin, who spent a decade as an estimator and salesperson, said it also removes an awkward moment where the client, excited and showing things off, waits while the estimator stops to write it all down.

Then comes the bid. A Utah landscaping client of Sanchez’s learned late about a request for proposal involving a three-year snow contract with a possible two-year extension across thousands of church properties. The client wanted to bid on 1,000 of them in two weeks, with one estimator. AI built the kickoff data and the structure for pricing readiness fees and per-push fees.

“I think it is moving a little bit more to like a shotgun approach and then refine afterward,” Sanchez said.

In its AI-compiled bid, the team only missed a couple hundred square feet of pavement. “But I think it will come out in the wash,” he said.

Keeping the human touch, but supercharged with AI

The price still belonged to the estimator. “AI, to me, is not making the decisions on an estimate. It’s actually helping the estimator make the decisions,” Brasington said.  

After the work is done, Sanchez said, comes what he calls “a feedback loop of estimating” —how the job performed against the bid, and whether the numbers change the next one, such as an installation rate for hedging. Brasington said AI can see every past job instantly, not only the one just finished.

Trusting the answers took the most care. Connecting AI to a business system is “like having a college student that just graduated,” Brasington said, so it needs guardrails. One of his clients asked AI to pull vehicle prices from Kelley Blue Book. The numbers came back wrong, and the AI later said it could not connect to Kelley Blue Book and had used the internet.

“No, don’t believe everything it tells you,” Brasington said.