AI in Building and Design: How Firms Are Using It Now

Central Florida design and remodeling firm KBF Design Gallery has been experimenting with AI for the past few years and is now actively integrating it companywide with success.

The firm is currently developing a proprietary solution using ChatGPT agents: AI-powered agents/assistants that can make autonomous decisions and complete tasks on your behalf, rather than just respond to prompts.  

In KBF's case, the AI agent doesn't just answer questions about project schedules; it actively monitors supplier pricing and market data to identify material cost saving opportunities across projects.

The firm’s initial AI adoption came out of necessity. When faced with staff turnover, brother-and-sister owners Adam Vellequette and Ashley Sheaffer realized their complex design-to-build process, managed entirely in a spreadsheet, was becoming unreliable. Changes to dates or details could cascade unpredictably, making the sheet increasingly unsustainable.

“We were consistently finding ourselves saying, ‘Okay, that date did not pan out. That shifted. This information is unreliable,'” Sheaffer says.

The duo started by recording conversations about the team, their capabilities, and task details related to the spreadsheet, gathering project timeline information, and then uploading that information into ChatGPT. 

From there, the AI could identify what was dependent on starting each plan set to stay on track and complete projects on time.

“Where before we were only able to look out a couple of weeks, we were now looking ahead several months," says Vellequette. "It was identifying bottlenecks that we were going to have because we were going to have three projects all hit a speciality at the same time, and we wouldn’t have enough manpower for it." 

With that information, they could then determine when to hire a new person and what role was needed based on project profitability. “We couldn't afford to have this bottleneck and we needed to hire," says Vellequette. "Where we would have been reactively hiring, we were [now] proactively hiring.” 

Product specification is one of the most promising areas for AI, according to 2025 AIA President Evelyn Lee. 

“It is a time-intensive part of practice,” she said. “AI can analyze and compare products based on sustainability criteria, performance metrics, cost ranges, and other project requirements. This helps architects move through the specification process with greater speed and confidence while reducing the likelihood of errors that can surface later in a project.”

KBF is applying that same logic to material purchasing. The firm has given an AI agent access to search for the best price for specified products, automating what can become a time-consuming manual task (and discovering cost savings in the process).  

The agentic AI logs into KBF's vendor portal, searches by model number, and documents the baseline price, Vellequette says. Then it crawls the internet to find that same model at a reduced rate, whether due to seasonal sales, overstock inventory, or other pricing differences. 

What might take a human hours of comparison shopping, this workflow delivers within 30 to 45 minutes. “We’ve had tremendous success with it finding some prices that are better than our vendor price," says Vellequette. "And that’s straight to our bottom line."

While KBF uses AI to streamline internal operations, New York City-based Kligerman Architecture & Design is exploring its potential on the client-facing side.

The firm is known for its high-touch, personal approach but has found ways to extend its work with the use of AI: structures imagined using only a pen and paper communicated through quick sketches (a signature touch), can now be rendered directly from the sketch as a watercolor, or a photorealistic view of the building, using the latest technologies.

“We are experimenting with the sketches we do by hand and taking them to the computer and exploring things such as material options, color options, or roof types,” says Joseph Carline, partner at Kligerman. “It can take seconds to rapidly visualize, and we can iterate faster than revisions of hand drawings.”

Not wanting to dilute the unique handmade nature of their work, the Kligerman team always starts with very well-developed sketches featuring their original ideas, and adding AI, layering in parameters within a prompt (thoughts, requests, caution against changing the geometry, etc.). 

“This gets us to a point where we can present a watercolor or sketch of a building faster,” Carline says, helping provide the client with more design options in less time to get to a yes. “It’s a tool, not an idea generator, and clients respect that.” 

One example he cited: adding elements like landscaping, fencing, or a pool—details that create a sense of place early in the process before a landscape designer is brought in. 

AI can accelerate or refine these design steps, but AIA's Lee agrees with Carline that human judgment remains essential to guide the conversation with clients, clarify what's aspirational and ensure the ideas work in the real world. 

“AI can support creativity and efficiency, but it cannot replace the judgment, intuition, and responsibility that licensed professionals bring to custom residential projects," she says. "This is where safety, durability, and performance enter the conversation. They test assumptions, consider structural implications, and make sure the design complies with codes and standards. They translate possibility into practicality.”