How AI Cuts Noncompliance and Cost Blowouts in Construction Projects

If you’ve ever watched a project drift off-track, you’ll know it rarely happens because of one dramatic failure. It’s usually a slow leak. A missed requirement. An assumption. An interpretation that doesn’t match the intent. And before you know it, you’re knee-deep in RFIs, rework, delays, and awkward conversations about cost.

In this conversation on Built. Trusted. Chosen., I sat down with Dennis Dixon (OQULi, creators of CODiii) to unpack a big idea that’s hiding in plain sight: noncompliance is one of the biggest silent drivers behind projects going over budget and over schedule. We also got practical about AI, what it’s good for right now, and how construction leaders can use it without losing the human edge that makes great work… great.

Trades were early adopters of AI on-site (because it saves time)

Dennis shared something that surprised me: on sites in Canada, trades and site leaders were among the first people he saw using AI in real life. Not for flashy tech demos. For simple, practical reasons.

Site supers were already chatting to ChatGPT because they didn’t want to type long messages on their phones. That’s the point: adoption doesn’t start with strategy decks. It starts where friction is highest. If you’re leading a construction business, that’s a cue to pay attention. The best tools win because they remove friction from everyday work, not because they sound impressive.

AI is most valuable as an assistant, not the “doer”

There’s a lot of noise about AI replacing jobs. Dennis’ take was grounded: the big wave right now is assist, not do.

That matters because it changes how you roll this out in a team. You don’t need AI to “take over” your business. You need it to help your people move faster, make fewer errors, and get back more time for the parts of the job that actually require experience, judgement, and leadership.

Dennis gave a great example: AI can help break design blocks. If an architect is stuck staring at a landscaping area for hours, AI can help generate options faster, so the human can refine and choose. Same principle applies in construction: use AI to get to “good options” quicker, then apply expertise.

Noncompliance is the thread behind blown budgets and schedules

Dennis and his team spent years listening across the industry, not just one corner of it. They spoke with all sorts of players, including suppliers and specialist trades, and they kept finding the same root issue: noncompliance.

That’s why he calls their product a “design and construction brain”. The goal isn’t to pretend one person can know everything. It’s to connect teams to the requirements and realities of a project so they can spot gaps earlier, and ask the right questions when they don’t even know what the right questions are yet.

Hundreds of handoffs create “interpretation risk” (and RFIs thrive there)

One moment that stuck with me: Dennis described showing multiple versions of the same footing or foundation detail, from different companies, all for the same application. Same intent… totally different execution.

That’s “interpretation risk” in action. Every handoff is an opportunity for drift. And drift is where RFIs multiply. Even when you get close to “one version of truth”, people will still interpret it differently unless the process makes it easy to confirm the requirement and catch the mismatch early.

This is why reducing errors isn’t just about “better people”. It’s about building better systems that make the correct path easier than the incorrect one.

Many industries skipped automation and jumped straight to AI

Dennis made a sharp point: a lot of industries skipped over a foundational skillset, which is automation.

Instead of building automated workflows first, many teams jumped directly into AI and now feel stuck, because they don’t know how to connect the pieces. If you’re a business owner, this is actually good news. It means the competitive edge isn’t magic. It’s learnable: start with repeatable workflows, automate what can be automated, then layer AI on top where it genuinely helps.

Trust is dropping, so leadership and reputation matter more than ever

We also talked about trust. Dennis’ view: trust is getting worse, not better. Companies are trying to block their IP from being scraped. Teams don’t always want full transparency because they fear it’ll derail delivery. And in competitive tenders, everyone is guarding their edge.

My takeaway is simple: in a world where tools can generate infinite “content” and infinite “answers”, the leader who earns trust wins. Use tech to speed up the work and reduce errors, then reinvest that time into clearer communication, stronger relationships, and better guidance.

Next step

If you’re leading a trades or construction business, don’t get distracted by the hype. Focus on what moves the needle: reduce noncompliance, reduce interpretation risk, and use AI as an assistant to help your team ask better questions faster.

Dennis Dixon is the Co-founder & CEO of OQULi (Canada) and the creator of CODiii, a platform designed to help AECO teams find compliance gaps, ask better questions, and reduce costly project friction.

Want to be a guest onBuilt. Trusted. Chosen.?