CFSEI Expo Keynote Urges Cold-Formed Steel Engineers to Lead the AI Transition

RALEIGH, N.C. — At the 2025 Cold-Formed Steel Engineers Institute’s 2025 CFSEI Expo, May 19–21, structural engineers gathered not only to explore innovations in cold-formed steel (CFS) framing, but also to confront a more sweeping disruption — artificial intelligence.

The keynote speaker, John-Michael Wong, PhD, S.E., delivered a thought-provoking and pragmatic address titled “Strategies for Integrating AI into the Structural Engineering Profession.” Wong, an associate at KPFF Consulting Engineers and project manager for the Innovation in Structural Engineering Grant Team on Artificial Intelligence from the National Council of Structural Engineers Associations (NCSEA) Foundation, brought clarity and conviction to a topic that’s equal parts exciting and intimidating.

“AI is not just some mystical thing,” Wong said early in his presentation. “It’s math on steroids — and a lot of steroids, and a lot of electricity, and a lot of caffeine.”

John-Michael Wong, PhD, S.E., kicks off his keynote address on AI and structural engineering at the 2025 CFSEI Expo in Raleigh.

From that moment, it was clear the keynote would be far from abstract. Wong guided the audience through specific, actionable strategies for embracing AI — not just conceptually, but practically and with urgency.

‘Exactly What We Need to Study’

Wong opened with a sobering reminder: structural engineering ranks among the top professions poised for disruption by AI and automation, according to a 2023 Goldman Sachs study. Rather than take a defensive stance, the NCSEA Foundation chose to act.

“One of the comments that was made is that ‘we need to kill AI before it kills us,’” Wong recalled. “And the reaction of the board was, okay, this is exactly what we need to study.”

The result was the formation of a national team — engineers from across the United States — tasked with building resources, providing education and developing a strategic roadmap for AI integration in structural engineering. The vision, Wong said, is to “embrace AI, to revolutionize and empower structural engineers to be leaders in responsibly shaping the future of the built environment.”

1. Learn about AI

Wong structured his keynote around four core objectives: learn about the technology, develop an AI policy, establish strategy goals and take next steps toward adoption. At every stage, he emphasized that the burden — and opportunity — rests with individual engineers and firms.

• NCSEA’s sample AI policy and getting started guide

He cited the California State University system’s initiative to provide AI training to 460,000 students and 63,000 faculty and staff. “The new engineers that you will be hiring,” he said, “are going to be coming into the industry expecting AI to be there.”

That expectation doesn’t mean replacing human judgment. “Structural engineers who use AI will replace structural engineers who don’t,” Wong said, echoing a sentiment shared by other professions in the face of similar technological disruption. AI, he argued, should free engineers to focus on their highest-value tasks: creativity, empathy and design judgment.

A slide from Wong’s keynote presentation answers the provocative question: Will AI replace structural engineers?

CFSEI Expo engineers listen intently as Wong outlines AI’s transformative potential in engineering workflows.

To demystify AI, Wong walked the audience through how machine learning models work, starting with a basic linear regression. He described how large language models operate on vector math — calculations familiar to any engineer who passed linear algebra.

“AI uses a lot of data in order to train it,” he said. “And depending on their training, you will get different answers.” AI models are not static — they learn, adapt and improve. “The AI systems you see out there now … they are the worst they will ever be.”

Engineers, Wong insisted, should understand these systems at a foundational level — not to become programmers, but to remain in “responsible charge” of the decisions AI helps them make.

2. Develop an AI Policy

Much of Wong’s message centered around one essential tool: a firm’s AI policy. “A robust AI policy, I would say, is the cornerstone for responsibly starting your company’s journey with AI,” he said.

He cautioned against using free AI tools without fully understanding how client data might be used or exposed. “If the product is free, maybe you are the product,” he said with emphasis, drawing attention to the risks of using free individual accounts with consumer-facing tools like ChatGPT.

Microsoft 365 users, he noted, have access to enterprise data protections — a potentially safer environment for AI experimentation. The NCSEA Foundation provides sample policies and frameworks for firms to use, including templates for internal access, risk identification and ethics guidelines.

Wong stressed that defining both “approved uses” and “restricted uses” is critical. “If you don’t define what is okay, you may stifle people. But if you don’t define what is to be avoided, you run the risk of business and data exposure.”

3. Establish AI Strategy Goals

One of Wong’s most memorable case studies came from his own office. Faced with a recurring challenge — identifying and extracting concrete mix design data from bundled PDF submittals — he trained Microsoft Azure’s document AI platform to do the task.

“Instead of doing it 10 times and doing it another 10 times, do it 10 times to train an AI model so that the AI model can do it 1,000 times for you,” he said.

The result? The system processed 78 submittals from 29 projects in 12 minutes and extracted 171 concrete mix designs into a structured Excel file.

“I didn’t have the heart to ask an intern to do manual data entry all summer,” Wong said. “Now I have AI.”

From there, he had AI generate charts, cross-correlations and statistical analysis — all based on real Bay Area project data his firm had collected over 12 years.

Don’t Avoid AI, Understand It

Wong repeatedly returned to the concept of “responsible charge” — the idea that licensed engineers must maintain oversight of engineering decisions, no matter how advanced the software becomes.

“Are you exercising responsible charge over the way that your software, your tools, are being used and deployed in your office?” he asked. “If you’re plugging into an AI system and you’re getting drawings and you have no idea what it’s doing, and it’s just a black box — I don’t think that fits the definition of responsible charge.”

The solution isn’t to avoid AI, but to understand it deeply, use it strategically and integrate it with intention.

A wide view of the 2025 CFSEI Expo keynote audience as Wong delivers insights from the NCSEA Foundation’s AI research initiative.

Wong shared how he used AI to extract key data from concrete mix submittals — automating a once time-consuming task.

Evaluate AI Tools

To help engineers navigate a fast-growing AI marketplace, Wong provided a framework for evaluating tools:

• Does this tool represent a threat or an opportunity?
• Does this tool help you, or does it compete with you?
• Who is the tool’s intended audience — structural engineers or others?
• Is it a product or a platform?

He cited examples like BUILDCHECK AI, which detects drawing inconsistencies across disciplines. And STRU.AI, which simplifies structural design workflows through a conversational interface. Others, like AREN AI, TESTFIT and DAISY.AI, are already being used by developers and contractors — potentially bypassing engineers altogether.

“Some tools may be positioned to help you,” Wong said, “but are actually competing against you.”

• NCSEA’s recommended list of AI and automation applications

4. Take Next Steps toward AI Adoption

Wong didn’t limit his keynote to cautionary tales. He showcased dozens of tools — from parametric design platforms like HYPAR, to document review tools like ReThink Plans, to low-level coding solutions for data tracking and workflow automation. He emphasized that engineers don’t need to become coders to use these tools. Instead, they need curiosity, strategy and a willingness to experiment.

“If you can save 12 minutes a day, it adds up to 40 hours a year,” he said. “What can you do with that time?”

Whether it’s summarizing meetings, managing RFIs or reviewing code changes — Wong showed that AI can handle the process, while engineers handle the judgment.

• NCSEA’s recommended list of AI and machine learning training resources

As he begins the final section of his keynote, Wong challenges engineers to take their next steps with AI this year.

A Path Forward for CFS Framing

Throughout the keynote, Wong encouraged attendees to reflect on how AI might specifically benefit the CFS framing community. 

“What data in your workflow is valuable in the world of cold-formed steel?” he asked. “What kind of past project information do you think is valuable for you as a firm, or valuable for you as an industry?”

For a material like cold-formed steel — where code equations are complex and project configurations are increasingly digital — AI presents an opportunity to unlock time, improve coordination and support better decisions.

Wong closed by encouraging engineers to start small, implement smart policies and embrace AI not as a threat — but as a partner.

“Let the computer do what the computer is good at,” he said. “And let the engineer do what the engineer is good at.”

With that, he left the 2025 CFSEI Expo audience energized, informed and, most importantly, equipped to take their next step into the future of engineering.

Additional Resources

• 2025 CFSEI Expo’s “State of the Standards” Charts a Stable Path for Steel Framing
• Insights and Inspiration Take Center Stage at 2025 CFSEI Expo in Raleigh
• Larry Williams Honored as CFSEI’s 2025 Distinguished Service Award Recipient