Deploying agentic AI to scale operational excellence

The Challenge

Bernards is an employee-owned builder managing dozens of complex projects at once across sectors like healthcare, education and civic infrastructure. As the company grew, legacy systems and highly customized workflows made it difficult to maintain consistency, onboard new hires quickly and get a clear, reliable view of project risk across the business.

The Solution

Bernards standardized on Procore as a single platform for project management and project financials, creating consistent workflows and clean, structured data across every job. With that foundation in place, the company is now leveraging Procore AI to put their data into action, facilitating informed decision-making at scale.

The Results

• Codified decades of learning with Procore’s agentic AI
• Standardized workflows across projects, improving consistency in execution and forecasting
• Identified project risks through data-driven analysis
• Improved alignment between field teams and leadership through shared data
• Built a foundation for AI-driven insights across project and financial data

Too much customization, not enough usable data

Michael Turl has spent nearly three decades at Bernards, starting in the field and working his way up through project management before moving into his current role as Chief Data Officer.

Over that time, he’s seen multiple waves of construction technology come and go. The lesson he’s taken from that experience is simple: tools don’t mean much if they don’t make the business more consistent. “We strive to be data-driven and data-dependent,” Turl said. “It’s about leveraging data to make better business decisions.”

For Bernards, that meant rethinking how it approached technology altogether. The company’s previous systems were powerful but highly customized. Over time, those customizations had the opposite of their intended effect, making it harder rather than easier to operate. “We had created our own way of doing things,” Turl said. “It worked, but it took too long to train people.”

That became a problem as the company grew. Bringing on new hires meant teaching them how to navigate a system that existed only inside Bernards. “We don’t want to spend three months trying to train somebody,” he said. “We want instant offense. When we hire somebody, we want to put them in front of the trade’s gold standard tool, and that’s Procore.”

Standardizing the way work gets done

When adopting Procore, Bernards chose not to recreate its internal processes inside the platform, aligning instead with how Procore was designed to work. “We didn’t want to break Procore by creating a bunch of ‘Bernards-isms,’” Turl said.

The result was a more consistent way of working across the organization. Project teams follow the same workflows, use the same tools and generate data in the same structure, regardless of the job. “If we have consistent data and consistent workflow, we believe we’ll have more consistent results,” he said.

That consistency has streamlined day-to-day operations, but it also allows helps leadership to compare projects, spot trends, and have greater confidence in the data they’re seeing. “We want consistency in our numbers, both from a day-to-day project management standpoint and from a forecasting standpoint,” Turl said.

Most importantly, that standardization gave Bernards a trove of data that could be understood and processed by the emerging generation of AI tools.  

Freeing teams to focus on higher-value work

Today, Turl is focused on how technology can change the nature of the work itself. That’s why it originally adopted Datagrid, the agentic AI solution recently acquired by Procore. 

For Bernards, Datagrid stood out for its ability to pull information from a wide variety of sources, streamlining how data is aggregated. “What differentiated Datagrid was the breadth of data it could ingest,” said Turl. “It can pull from Procore, Textura and Vista, allowing us to see trends across our entire enterprise data lake.”

Though the company is still in the early stages of its work with Datagrid, it’s already applying the tool to time-consuming administrative tasks. “The idea of paying talented and highly qualified people to do clerical work is unsettling in the emerging age of AI,” Turl said.

Instead of digging through specifications or assembling reports, teams can use AI tools to surface the information they need and focus on interpreting it. “It’s about removing the tedious work so people can do more thought work, and I really believe AI is the key,” Turl said.

“We can create an agent that acts as a one-to-one match for an SOP,” Turl continued. “As it finds data outliers, it creates a teaching moment for the individual — it doesn't just fix the data; it teaches the person what good looks like.”

That shift is already starting to take shape in areas like RFI research and submittal reviews, where AI can help identify relevant information and highlight potential issues. “We get 'quick offense' by leveraging Procore AI's large language models for RFI research and submittal reviews.” Turl said. “What used to take a professional an hour or two to flip through a spec book, AI can now do in seconds.”

Leadership is even looking to use Procore AI to create pay packages, combining data from sources like Egnyte and sub-billings, so they could process payments in Vista Viewpoint. “That would cut 60% to 70% of the busy work in accounts payable,” said Turl. “That’s potentially the equivalent of two or three full-time employees.  Those individuals would be able to refocus and add value to other elements of the business.”

The long-term application, however, goes beyond administrative tasks. “The unquantifiable component is what challenges our employee-owners are able to solve ahead of them being a challenge,” he said.

Scaling without adding complexity

As Bernards continues to grow, Turl’s focus is on increasing the scale of the business without increasing its complexity. That means continuing to invest in systems that standardize how work gets done, improve the quality of data and drive better decision-making across the business.

It also means building on the foundation that Procore provides. By combining structured data, consistent workflows and emerging AI capabilities, Bernards is working toward a model where risks are identified and teams can operate more effectively across projects. All supported by a partner that’s as invested in the industry’s future as Bernards and Turl are. 

“The acquisition of Datagrid makes me 100% more excited to partner with Procore for the foreseeable future,” said Turl. “We value that Procore is deeply invested in our success and open to true partnership in product development."