Writing the Book on AI for Residential Construction
AI is taking off across the housing industry, and Grace Mase, an architect, contractor, and author of AI in Residential Construction is encouraging builders to get on board. In her book, Mase outlines what AI can do for residential builders and provides resources to help improve nearly every aspect of business.
We invited Mase onto the Women at WIRC podcast to share some of those insights. Listen to the full episode:
Transcript:
Welcome back to Women at WIRC, where our editors from sister media brands—Pro Builder, Pro Remodeler, and Custom Builder—sit down with standout women across home building, remodeling, and design. We share their stories and business insights, and explore how women are reshaping the residential building industry.
Rich Binsacca: Hello everyone, and welcome back to the Women At WIRC podcast. I'm Rich, from Pro Builder, and your host for this episode. Today I'm joined by Grace Mase, an architect, licensed contractor, and the author of AI for Residential Construction, which was published earlier this year and available from NAHB’s builderbooks.com.
The book combines Grace's vast and varied housing industry experience with a track record of more than 20 years in machine learning, what we now call artificial intelligence, to give builders a gateway to applying AI to their operations and specifically production processes. To achieve greater efficiencies and profitability.
She's also allowed us to excerpt parts of her book for a feature article in the March/April issue of Pro Builder entitled The AI and Labor Connection, which you can find and read at probuilder.com.
Grace, you wrote this [book] for residential construction. I think it's an awesome book.
It's published out of NAHB’s builderbooks.com. Give me a little background, how you got started, your interest and relationship with the home building industry or construction industry. Just give us that little background so we understand a bit better.
Grace Mase: Sure. Thank you so much, Rich. I'm a daughter of an electrician. When I was 5 years old, my parents saved enough money to buy their first home, and that really is the beginning of my love for construction, especially in residential. My dad would take me to the job site every Sunday morning after church, and would just inspect the work.
And it was so much fun to see every aspect of construction. And I remember one particular morning my dad was standing there in the sunlight, his face at just the right angle, and his face lit up with a massive smile and a sense of pride. I would never forget he was officially a homeowner and the thought that he was able to provide a safe and comfortable home for us—that meant the world to him.
And the sense of pride and security has stayed with me. So when the time came around to study, you know, to pursue the construction industry, at that point, it seemed like architecture was the only path. So I went to Berkeley, got my undergrad, and afterwards, that period of time, I experienced the Loma Pre Earthquake and Oakland Hill fires within a span of two years.
And then afterwards I worked as a campus architect focused on seismic retrofits and fire safety. So there's a strong correlation there.
I earned my Master’s degree in architecture at Yale and I learned to design and build a ground-up home with Habitat For Humanity of Greater New Haven. And so I also developed, during that time, I also developed Yale School of Architecture's first website.
So, since then, I've been building both the physical world and digital world. And today, I hold multiple residential or contractor licenses in multiple states. So that's the beginning of my passion for construction. What drove me, inspired me to write the book was like many families downhill Los Angeles, we had to evacuate from the LA wildfire, and our friend's home didn't survive, and so I helped them to rebuild, and I saw many contractors continue to misinterpret the like for like as build what used to be, which is a missed opportunity because if they continue to build what they used to be, they'll burn down again. We know the building science exists, better building materials exist, and incentive to reward them to build better exists, but we're not having those conversations.
And that's what I realized. AI can help us to get that information at the palm of every contractor so they can actually build better and earn more on top of that and deliver a better result for their homeowners.
To me, I think what happened to Oakland, what happened in LA are significant. But the wildfire, I mean, even LA just looking back this year, when they had time to start assessing what the damages were, $164 billion in property and capital losses. That is huge.
Rich: Right.
Grace: 80% of homes were built before the modern resilience standards were common.
That means four out of five, and not just here in Los Angeles or Oakland. This is nationwide. Four out of five homes would not pass today's building code. And that's a skill and vulnerability we're talking about. And for us as a construction industry, we need to start thinking about how do we upskill so we can actually make a difference?
That's a huge volume we're talking about, and that's the greatest opportunity our industry will ever face. But at the same time, we need to sharpen our tools and figure out how to do so. And AI itself is not just one tool. There are multiple AI tools to do different tasks, and that's the part I'm super excited about, and I want that information to be accessible by every contractor.
Rich: So, we talked about what inspired the book, and I think that's a great inspiration or great impetus for kind of saying, this needs to be brought out into the open. We need to get this in, like you said, to the palm of the hands of builders and contractors who are working on new homes, and retrofits, any construction site, really was the primary message you wanted to get across. Because to me, what we've covered in AI and, and what we've, when we've surveyed our audiences and have seen surveys of builders through NAHB and other folks is that there's interest but not a lot of implementation. How did that work into how you thought about messaging this book and the messages you wanted to get across?
Grace: So for me, I've been building AI software for the last two decades or so, and what's brilliant about AI is it could be your personal research assistant that's personally to you, geographically located just for you and help you find money that you didn't know you had.
And also help you to, you know, become or have a communication coach to help you to frame conversations so your point will land better with your clients. So reduce the friction between the, you know, your relationship with your clients. And the other part is also great because it could actually translate information.
Perhaps some of your clients may not be proficient in English and having an AI tool to do the proper translation for you so they can deliver exactly the scope you want them to deliver. And that's a game changer in our industry. And the other dimension that doesn't get talked about is construction has one of the highest percentage of neurodiverse people by means of people with ADHD or dyslexia. I think the UK did a study where 52% of the construction industry are neurodivergent. So that’s one out of two, that's a pretty significant impact.
Rich: Yeah. I didn't realize that. That's interesting. And I can see where AI, again, kind of simplifying or translating into a narrative or into instructions that are more easily understood, more easily followed, could really be a help. I didn't know that stat. That's very interesting. Okay. Talk to me about the internal or external factors that are driving builders to AI.
It is, obviously, it's some of the stuff you're talking about, but in your experience and when you talk to builders, and you reference builders in the book, what kind of got them to say, ‘I need this’? Was it a particular pinch point? Was it, ‘I need to get ahead of this’? What did you find? I'm sure it's all over the map, but what got them into it?
Grace: I think if you look holistically at the entire process, there's friction along the way. There's not just one. There are multiple issues, such as labor shortage, skillsets, rising material costs, margin pressure, client expectations to limited internal resources, to engage and also the ongoing, growing changes in our regulations, is complex.
And if we start reducing now and using AI to help us do better on addressing some of these issues, not only will we improve our productivity, we'll also improve accuracy and make better decisions. And we started thinking about all three. Those compound into real margin improvements. And oftentimes when I think about research, even the compliance or materials, the number of hours we spend and dedicated resource to do so, we're talking about 48 to 72 hours per project to do that kind of research, we're talking about $4 to $5 to $7,000 of savings just on one project, just the front end of pre-construction.
If we can do all that better and have a system to even cross-reference multiple documents of, you know, for architects to engineers and identify issues that we can address before signing the contract with our clients. That means less change orders, less frustrations, less delay and permitting or addendums that need to be issued.
We're saving thousands of dollars. I mean we, when you look at holistic processes, we're bleeding money left and right. If we stop reducing those gaps or reduce those frictions, we're literally gaining higher margins. It's that simple.
Rich: I think that speaks to the attainability issues that we're having, that speaks to waste reduction. It speaks to profitability, obviously. I think that's, you know, all things that should resonate with builders and should inspire them to say, ‘I need to look into this and figure out how I can leverage this especially. And, what I loved about the book was it was focused on construction issues. Obviously, you talk about different applications of AI, and there's a lot of them for the industry. But that's what intrigued me the most was that I hadn't heard a lot of talk about how AI can make the production process more efficient. It was more about sales and marketing or, you know, construction documentation.
I didn't see a lot of, ‘Hey, on the job site, this is where it's applicable’. So, the book does a really great job of addressing that specifically and deeply. and it seems like that's a, that's, as we know, that's where a lot of cost is for builders. Like you mentioned, a lot of waste that could be taken out of it, that directly affects cost, pricing, then profitability, certainly. So, what's the problem here? Why aren't builders getting into this? What's the barriers that you found? What's the hesitancy? Do they just not know what they don't know? Or is it something else?
Grace: I think you hit it right at the last one. Fear of the unknown.
They don't know what they don't know, and the thought of, ‘I had to invest a lot of time to figure it out,’ because technology unfortunately may not be that intuitive in some of the applications. And that's disappointing, and that's an opportunity for the industry to do better with construction tech.
And so that to me is a tremendous opportunity if we can figure out ways for builders to learn the basic baby steps, if you will. And that was the genesis of writing this book. How do I even show them how easy it is to use and just have a conversation with AI to learn more and figure things out? That could demystify a lot of things for most of us, you know, when we don't know what we don't know.
Just like clients when they don't know what they don't know, their immediate reaction is, ‘I'm just going to stay with what I know,’ or if things don't work out, of course it's someone else's fault, right? There's that extreme cases. But if we just slowly progress, disclose the right information for everyone, all the builders, to try just one tool and get the little quick wins under our belts, then we feel better and more confident trying out bigger things.
Just start little; you don't have to overcome the entire AI system or start building up your MCP servers or avail up your agents. But let's just start with very basic stuff and learn some of the basic tools. Get comfortable with it. Ask those critical questions that you may not understand and start engaging and getting dialogues and getting more insights.
That's how you ladder up and when you ladder up, then you'll be comfortable taking out the next set of tools to explore.
And I also recognize that everyone's so busy. But if you can carve out 10, 15, even half an hour on Friday afternoon just for yourself, or even have it to do with your whole team, so everyone ladders up together, that's a great opportunity.
Rich: Right. That's a great way to put it. And I think that builders are time crunched. They feel like, you know, how am I going to find time? It's a big animal. How am I going to get this done? And so starting small and starting. You know, with a small amount of time, relatively, and just, being dedicated to that.
And I like the idea of kind of getting the team involved at the same time because then, like you said, everybody ladders up at the same time and you're not just having one person who knows it, the whole team or everybody who needs to be involved in that process is aware and learning at the same time and helping each other to get there.
Do you think, is there any validity to thinking that it somewhat depends on the size of the builder, the resources that they have in-house, the resources they have to find somebody out of house. It doesn't seem like from your description that it really matters, but have you found that? Because we've found that larger builders, builders who have internal resources that recognize the value or at least can have the latitude to learn the value and how to use AI, or adopting it more often or in greater numbers. Any perspective on that?
Grace: I don't think size matters. However, I do see the mindset.
And oftentimes if you have someone in the team that really finds value and is able to prove that there's actually cost savings or a time reduction. And that's actually a really fun thing to see that we need to treat AI as how we learn to use the tools.
Little by little, try it out to see that and track the performance. How are we able to reduce time and reduce costs for everyone? And ultimately, how are we helping the company to grow and how are we delivering results for our clients? And those are the things that matter.
Rich: I think what's great about the book, too, is that you provide resources.
Like if you're looking to do this, here's some resources you can go to. If it's this, here's some other ones you can find. And I think the other thing is that, ‘Where do I start? Where do I go looking when I've got this problem and I want to solve it? It seems like AI will solve it or help me solve it, but I don't know where to go from there’, and I think there are resources out there. Obviously you've, you've pointed to a lot of them in the book. And I think that, that's a tremendous resource, part of the book is that you give them the place to start depending on what they're looking to do.
We talk about data and we talk about AI. I think one perception that I've heard builders and others say is that, well, AI just scours the internet and who knows if that's good information or bad, or can you trust it? And I think what's missing, and I want you to kind of set me straight if I'm wrong about this, but when we're talking about AI for a company, you're drawing data from that company, you're drawing data from what they have available and what they're gathering. So, it's internal. It's not necessarily external, although you mentioned some things, like looking at weather patterns and things like that, that are external but are available, and can be brought into that intelligence.
But, really, you know. It's internal data. Am I correct in saying that for the most part, the AI's drawing from internal data?
Grace: Well, so there are two parts. Your internal beta data is key. AI can also, they are trained, many of these large language model AI tools are trained with data that's available in the ether.
And so because there's a vast amount of data. When they synthesize and analyze the data, most of the time it's pretty close. For example, estimation, you already have your estimate. You know how you can declare like $300 per square foot depends on parts of the country, right? And you already have a good sense how much it can cost you from demolition down to you know, finishes.
In reality, the large language model, many of them are being trained with a lot of data that's available. And when you start comparing, some are pretty close. And so that's encouraging. And then that comes down to the next step of how long does it take you to generate an estimate on your own versus leveraging AI tools.
That's gathering data from external sources. Or you can use it for internal as well. You can, there are different ways to set it up and be able to generate your estimate. Minutes versus hours. That's a cost difference and that's a missed opportunity, right? If you spend hours to generate an estimate versus minutes to generate an estimate and have a few minutes to scan through, make sure that mostly is accurate.
Think about the amount of time you can do something else. That's opportunity cost. Imagine if you're able to save hours. On your daily operation to do something that's more impactful to grow your business, connecting with your clients. Those are the ones that actually will have paid dividends down the road.
Rich: Right, and going back to the estimate, it's like how often does a builder or a contractor kind of audit their estimates? Or am I being accurate here? Is it really the best cost that I'm using, and at the very least AI and AI tools could basically say, ‘Well, this is what we found for that same situation’.
How does it compare, and you can kind of go, ‘Oh, okay, there's a sense check there. Maybe I need to kind of hone my estimates’. And then at some point, having enough data from you and the world. that it's been trained on to be more accurate, to be as accurate as possible, and do it in, like you said, just a fraction of the time, and allowing you to do other stuff that's more impactful and more customer facing.
Grace: Imagine the time you have, that extra hour, let's say. If you're able to start doing research and leveraging AI to do material research, figure out which one's available, which one's more affordable, which one's actually better quality, and what the lifespan for the materials, and do all that analysis and synthesize all that information and share that with your client.
Imagine that conversation. They will be impressed. They realize that you are looking out for them. That trust is built immediately. That's valuable.
So, that's the part I get excited about. I mean, I think there's also a mentality of, you know, garbage in, garbage out. And people worry about if their data is messy. It's like, there may be, but reality is you can also leverage AI to sanitize and clean up your dataset. So you can create the structure that you need to do further analysis and all the performance that you may need. Just like, here's what was my estimate, here's the actual, and what's the delta between the two?
And how do I do some predictive modeling, figure out what do I need to do better next time? And that's intelligence. Imagine you have that now and next estimate put together, you can be much more profitable
Rich: In your experience and talking to builders and not just the builders that you talk to for the book or have experience with in the book, but, just being in the industry and kind of knowing a lot of builders, are there particular pinch points and specifically on the production side of it, the construction side of it—maybe we've already alluded to it or spoken to it—but I want to make sure people understand that these are some of the pinch points that builders have and it might be something that they can relate to and say, ‘Well boy, they solved it with AI or helped solve it with AI. I have that problem too’.
Grace: I think we touched upon it a little bit, but there is that co-research, the updates, the incentive plans, the material research, the estimation, scheduling, client communication, all those are real friction points. And these pinch points often lead to longer delays and resources dedicated to resolving them.
Imagine, to be able to do a quick lookup, cross-referencing, make sure everything's done correctly. AI is phenomenal doing this kind of repetitive research, heavy lifting, and by doing, getting those done quickly, you actually get things done better. When things are done better, the clients see us as, hey, you can resolve these quickly.
You figure things out and get all the data I need to make my decisions, smarter decisions, it's easy conversation, and then be able to help our clients to understand holistically, the big picture, the full picture really, of the value chain associated with that.
Not only are the incentives, grants, tax credits, and all the good stuff, rebates are available for us. There's also utility cost reductions. If you start building out better quality, better equipment, that's also significant. We're talking about thousands of dollars a year, right? This for us, I think most of us will appreciate that. And also insurance premium reduction. That's also pretty significant.
And for us to figure out which building materials actually have the same quality or structural integrity, but still be able to perform at that level but at a reduced cost when we want to know that. And if we can have that value conversation with our clients, then we no longer focus on what's the lowest bid and we will feel better about delivering something that's valuable for the client.
And we'll be proud of our work, too.
Rich: Everybody seems to win in that scenario and I see that point. So AI's moving, it seems to be moving very quickly. You've been involved in it for 20 years. I think people would be surprised because they think it all started in 2023 and it didn't, believe it or not.
But I mean, it seems to be moving very quickly. Data centers being built, electrical grids being stressed. A lot of money is pouring into that industry, for the construction industry. But maybe just generally, where do you see AI needing to or continuing to improve, and how is it improving?
Grace: AI is great, synthesized general information, but when it comes to construction-specific information, the precision matters to us, and so we need to make sure that the data is well trained and accurately represented and timely. More importantly, because most of these large models are trained with certain periods of time intervals, and so anything that's most recent may not get trained, and that’s information we have to be mindful of.
Look at other resources cross-referenced to get the accurate information until they get to the point where everything is real time for the time being. We need to find a way to resolve that issue. And then the other is about communication, supporting the vast majority of our population in construction, neurodivergence, and we need to make sure we allocate enough resources and bridge that communication gap.
Rich: So this is terrific. Again, I can't say enough about the book. I'm glad I found it. I'm glad that I got a chance to get it from NAHB. Again, it's available at, published by, and available through NAHB Builder Books, at builderbooks.com.
Anything else, Grace, that we haven't talked about on this topic? It's been terrific. I think it's been comprehensive. Is there anything else that I missed or that you want to make sure gets across?
Grace: Well, my ask to every builder or every remodeler is don't wait for your competitors to start using it. Start using it now, and the book does walk you through step by step how to get started.
It's really not that complicated, and we just need to have the same kind of curiosity as we do in the construction world. How do we solve a problem? How do we learn? How do we upscale? And, so I wish everyone well, and I wish everyone being prosperous. This is an exciting time for all of us in the industry. We can make a big impact not only for our clients but also specifically our communities.
Rich: To buy Grace's book AI For Residential Construction, go to builderbooks.com, and thanks for listening.
Thanks for listening to Women at WIRC. This podcast is a spinoff of our annual Women in Residential Construction Conference, which we’ve hosted since 2016. You can learn more about the conference and see when we’ll be in your area by visiting womensconstructionconference.com. Women at Work is a production of Endeavor Business Media, a division of Endeavor B2B. Until next time, keep up the good work.
Women at WIRC is produced by Custom Builder, ProBuilder, and ProRemodeler. Subscribe to the podcast at:
Or wherever you get your podcasts...