AI in Construction: From Basic Tools to Advanced Automation and Beyond

This week, we explore how AI is transforming the construction industry. Discover how AI tools can automate tasks, streamline project management, and improve efficiency, ultimately making your business more profitable. Whether you’re already using AI or just curious, this episode offers practical insights to help you leverage this powerful technology.

Topics we cover on this episode include: 

• 00:21 – The role of AI in enhancing productivity
• 01:18 – The evolution from simple prompts to AI assistance
• 01:56 – AI applications in estimating and bidding
• 02:50 – The role of AI in contract management
• 04:08 – Chained AI agents and their practical use
• 04:54 – Voice-activated AI and customer service
• 05:58 – AI’s role in risk management and contract analysis
• 09:08 – Automating document analysis and project specs
• 11:02 – Using AI for contract review and identifying risk
• 14:12 – Integrating AI tools for better project managemen
• 15:56 – Future possibilities and AI-driven efficiencies

LINKS

Join our free community for construction contractors

Join the conversation on our LinkedIn page: https://www.linkedin.com/company/CarpenterCPAs

FIND US ONLINE
Wade Carpenter, CPA, CGMA | CarpenterCPAs.com
Stephen Brown, Bonding Expert | SuretyAnswers.com

GET IN TOUCH

Questions? Want to find out more? Get in touch, we’re here to help! Fill out the form below to get started.

TRANSCRIPT

[00:00:00] Wade Carpenter: Have you ever thought about how far AI has come in just the last year? From simple text prompts to full fledged AI assistants, the technology has evolved rapidly, and it’s transforming the way we work in construction. Imagine if you could automate not just the mundane tasks, but also streamline your entire project management, bidding, and client communication processes.

Today, we’re exploring how AI is revolutionizing the construction industry, making businesses more efficient, accurate, and ultimately more profitable.

Welcome to the Contractor Success Forum. I’m Wade Carpenter with Carpenter Company CPAs. And as always, I’m joined by my co host, Stephen Brown with McDaniel Whitley bonding and insurance.

Stephen, AI is not just a buzzword anymore. It’s becoming an essential tool in the contractor toolkit.

[00:00:46] Stephen Brown: It’s a real deal. And I guess ever since I subscribed to Chat GPT, I use it all the time. And I love it. You were just saying earlier, now you’re using it every day. Before the podcast, I said, Wade, could you live without it now? At what point would you say that you couldn’t live without it?

Where are you on that scale?

[00:01:08] Wade Carpenter: It’s definitely made me so much more efficient and I guess if we didn’t have it, I would probably lose a lot of productivity personally. There’s a lot of things that I’ve gained in the last year and a half or so embracing AI as far as like researching things and as well as automating things.

And we recently did an episode on automation in construction, and part of this is a continuation of that. We’re talking about how this AI can help in just streamlining your operations and saving labor and actually making you so much more efficient and ability to go beyond a one man band type construction company. To scale up.

[00:01:49] Stephen Brown: Now that makes perfect sense.

[00:01:51] Wade Carpenter: The way I see it, the initial usefulness of ChatGPT, it was limited. We’d, you know, hey, okay tell me about this, or write up a email for me, respond to an email. And we would take it out of there, copy and paste it in, or type up a prompt, and we evolved from that into these things like custom GPTs, where we say this is what I want you to do, and this is a structured output that we want to get from it.

So we’ve gone to that part and it’s helped make us a lot more productive. And then we’ve transitioned from that to AI assistance. And now, we have things that are basically building your own apps today that it will actually do that for you. So some of the software that we may have been paying for in the past, we can actually build ourselves. And it’s amazing.

[00:02:43] Stephen Brown: Wow. So what are you using it for most right now?

[00:02:48] Wade Carpenter: It’s hard to say what I’m using it for most. We have several automations that we talked about last time that it will actually go out and, like we’ve got some data that we want to do something with or generate a marketing post or something like that. And we have a structured format and it’s automated and it’ll come back and write something up for us. And we all probably still need to look at what’s being put out, but we can humanize that content.

We’ve gone from things like these chatbots we can put on their website where they can not just, we had things like intercom or whatever, and you have these little chat bubbles in the bottom of your website, to where it would just say okay, ask a question, that would give a canned response that you knew was canned and you hated talking to these little bots, to now chatbots that will give you a custom answer based on knowledge that you give it, say about, what your services do and how you work and answering questions that a person would otherwise have to do.

[00:03:48] Stephen Brown: That makes sense. An English teacher friend of mine was talking about how mechanically the AIs are at writing things and what the kids are doing to turn in. But what the AI is not able to do, you can say, make it more human, make it funny, make it this. It doesn’t show your personality.

And he says, I know my students’ personality, and I know their abilities, and you can spot the AI a mile away. Mechanically, you can’t go and just terrorize them and circle everything in red ink, because mechanically it’s perfect.

[00:04:23] Wade Carpenter: I know that I’ve heard some cases here. I know one in particular here in Georgia where somebody used something like Grammarly or whatever. They have these AI checkers and they look for certain keywords. So, some of the things where don’t use these common phrases that AI always puts out there to humanize the content.

Other ways we’ve been using AI, and I’ve seen people using AI, it’s not just these chatbots, but the extension of that has been where we’ve actually brought voice into these conversations. And they’ve gotten so good that it’s more real time, and in certain cases you may not even know it’s a bot.

I personally would say you probably want to tell them that this is a chatbot that you’re talking to, but they’ve actually got it so that you could have a voice assistant answer your phone and route your calls. So you don’t necessarily have to pay a human being to do that.

Now, again, it’s getting better and better, but it’s amazing how this voice technology has taken off. So I need to slow down and let you talk, but I feel like I’m geeking out again.

[00:05:28] Stephen Brown: No, keep geeking, go ahead.

[00:05:30] Wade Carpenter: Just continuing the thought about these voice activated receptionists. And you think about, we have these booking links from Calendly or something like that. But you can have a voice assistant that will make a call and say, I’d like to book an appointment. And they have your calendar and actually can get the details on the phone without doing that. And you can have somebody answering your phone 24 7 that is a fraction of the cost of paying a human being to do that.

Today is more about considering the potential and some of the challenges of implementing some of these tools in a construction business, but I think one of the messages that I wanted to have is a lot of these tools are really coming into their own now so that you can actually start making some huge gains in your business if you lean into it. You got thoughts?

[00:06:20] Stephen Brown: My thought is you’re making gains on your business if you lean into it. That’s a great point. I think more specifically for our listeners in AI, it was so helpful when you were geeking out, Wade, because you were explaining to me and to our listeners, the evolution of AI and your use of it.

First thing that pops into my mind is at McDaniel Whitley, there’s always someone that answers the phone. And it’s always someone that probably knows you, and cares about you. Cares about you because we’re friends. We like those kinds of relationships. I don’t see AI replacing that so much, but I see the incredible, function of it in pushing sales.

I think AI, we talk about managing historical information for contractors, historical bid information. Worth its weight in gold to a contractor. What do you bid? What are other people bidding? What are they seeing that you’re not seeing? They’re always asking that question when they’re preparing a bid.

And, how much profit can I make on this job and have a chance of getting it? Because it costs money to bid work. It does. What are your thoughts on that?

[00:07:32] Wade Carpenter: Yeah, there’s some other things that I might bring up at another time, because I have seen some things like with the estimating and bidding, some things that will do takeoffs on blueprints that, manually done, could take hours and hours. And I actually saw a demonstration. I’ll make contact these people about maybe coming on to the podcast.

But they showed a demonstration, they had three experienced estimators given the same set of plans. And this one guy had done a takeoff on this blueprint in a fraction of the time, where, first guy was actually just still measuring the first floor.

What you mentioned about the bidding process, where we can build these scrapers are going out, your suppliers list their prices and you know what you need on that. You can go out and accumulate a lot of the prices and the information you need for the bids in certain cases.

But it could automate, like you’re a general contractor and this is what the bid is and we can say, spit this out to these subs. We want bids from these different subs. Give them the specifications and save so much time. There’s so many applications of that. I’m just feeling like I’m going down a rabbit hole here.

[00:08:42] Stephen Brown: No, that’s what I enjoy us talking about. The different applications of it. The things that we don’t think of. The things our listeners may not think of. I think about the whole concept of risk to a contractor. One way to alleviate risk is to do the things that should be simple perfectly over and over again.

What a great use of AI there.

[00:09:05] Wade Carpenter: AI is not 100 percent infallible.

[00:09:07] Stephen Brown: No, but it helps.

[00:09:09] Wade Carpenter: It is actually usually a lot more accurate than humans.

Chaining AI agents

[00:09:13] Wade Carpenter: There’s other things like chaining AI agents. So if you have a bot that does one particular task. And they can do that one particular task very well, and then it needs to do another part. One part of it may be accumulating information for the bid, one part may be putting it all together. And so chaining all these agents together is pretty amazing when you see some of these use cases.

Using AI for Research

[00:09:36] Wade Carpenter: Some of these other things, getting information, research. Perplexity AI, if you’ve never checked out Perplexity AI, that is giving Google a run for their money. And I think Google is trying to catch up on their AI technology. But if you’ve ever want to research something, go try Perplexity AI, even the free version.

It’s it will go down, it’s I want to know about this. It will start going through researching the topic and then answering your question. And you can see the progression of how it’s coming up with the answer, and it will give you sources of where it came from.

So you can check that out yourself.

[00:10:12] Stephen Brown: I will. That sounds like fun, Wade. I’m looking forward to that.

[00:10:15] Wade Carpenter: I use that one all the time. Even if you’re like researching buying something off Amazon or, what’s the best TV to buy? I don’t know what, whatever that might be, but–

[00:10:25] Stephen Brown: And it’s not going to nuke you with ads if you ask it that question?

[00:10:29] Wade Carpenter: No, it’s, you don’t see any ads because it’s going, you don’t have to sift through all that and,

[00:10:34] Stephen Brown: Was hoping that’s what you’d say because these days on a search engine You can’t ask or look at anything without being, nuked.

[00:10:42] Wade Carpenter: Right.

[00:10:43] Stephen Brown: Yeah. Okay.

[00:10:44] Wade Carpenter: It saves a ton of time on that kind of stuff.

But like I said, once you’ve done the research, I repurposing content like we talked about last time, this is where we were talking about building automations for a contractor. You got our podcast, and we take our transcript, and we will put it in to an automation that goes through different things, and we have a little bot that comes up with a title for a blog post, and then one that will create an outline for a new blog post based on the content.

Then another one that will write the segments of the blog posts and put it all together. And again, always want to have a human looking at it. But think about writing that from scratch. It saves so much time.

Part of the last episode was about tying all these apps together. You may have a proposal software that doesn’t talk to your CRM and all this stuff, and this automation where, maybe we do need to create some custom content that goes in our proposal. We have these no code tools, like we talked about, make. com and Zapier and some of these other ones that are really good, where they will use this API, which is Application Programming Interface, these integrations to allow you to push the data to one place, bring it back, and do a task without you having to manually copy, paste, and iterate through writing up your own proposal.

Does that make sense?

[00:12:09] Stephen Brown: Now it’s really going to help people, especially those that are putting together proposals for government contracts. Putting the best of the best in there and getting a rough draft knocked out relatively quickly that you can fine tune wi ll save you hours and help you to be even more focused, do a better job on your proposals.

[00:12:29] Wade Carpenter: Yeah, absolutely.

[00:12:31] Stephen Brown: I’ll have to confess, Wade. I wrote a letter for the Memphis Builders Exchange as a member of the board of directors, and when I presented it to the board members by email, one of the board members wife was an English teacher, she said that’s the perfect letter. And then I confessed to everybody that AI wrote it. And they were just fascinated with that.

And then as I look at a perfectly crafted AI, and then I want to make it personal. It doesn’t read the way I want it to read. Then I start messing around with it and then I have to get AI to correct my own grammar and punctuation.

 It’s a tool that you can use. It’s a tool that works. It’s fun to play around with. And it’s even more fun to think about how you can use it to be more productive because that just takes stress off your plate. A lot of stress.

[00:13:21] Wade Carpenter: You just brought up another use case that reminded me that I didn’t have on my list today. But one of the things that I’ve used, and our contractors I think should probably think about this, whether it’s a construction contract or whatever, you probably still want to have a good construction lawyer look at a contract, and I’m not advocating you not do that.

If you’re in Florida, you probably want to call our good friend, Alex Barthet, who’s been on the show several times. If you take that contract and you throw it in ChatGPT, I’ve done this with not just, give me the main details of this contract, but I’ve also said what are the cashflow implications of this thing? Where is it that I could be stepping on my own toes?

And that might be a really good step before you go talk to an attorney. It’s explain this to me. I know ChatGPT said this, and I would not rely specifically on legal advice from ChatGPT or any of these other models. But, it’s a great first step.

And we built tools where we’ll take just this PDF of like, a contractorr bought the truck. Let’s throw it in there and say, pull out the payment term so we can do the amortization schedule. But, when does this thing due and what are the down payment and what’s the interest rate?

And it can do a beautiful job of summarizing that kind of stuff.

[00:14:41] Stephen Brown: And that makes a lot of sense. I see looking at, specs, project manual that is a thousand pages and finding quickly what you want used to be by going to the table of contents. And then different owners, architects, engineers would have their own way of putting things together.

And so there’s nothing uniform to it. Now you look at that document, you go to the search engine and just from experience, what key words or phrases you need to pull up that section. I do that all the time now. I don’t know how I could operate without it. There’s got to be AI search engine that allows that to happen.

[00:15:21] Wade Carpenter: One of the things you bring up all the time is avoiding risk or, mitigating risk, recognizing risk. And different applications, but a contractor might use it in this particular context where you throw that thing in there. It’s like, how could I get hurt by this contract? Or, what are the potential pitfalls? Play the devil’s advocate.

And I’ve done that with different, not in the same context that a contractor might use it, but what do I need to know? One of the things that contractors have problems with cash flow, they don’t know the payment terms, and it may be buried in that contract. They don’t get their payment terms right, they turn it in this month they’re going to have to wait another 30 days. Specifying what’s, what are those things we need to know to get those kind of things done?

Another thing I’ve done with even just subscribing to a software tool or agreeing to terms and conditions on any website. I’ve copied them and pasted them in ChatGPT and Claude and some of these other models and saying what do I need to know about it? What am I signing up for? And it’s amazing what it’ll tell you.

[00:16:24] Stephen Brown: That’s great. And when you say hidden in the contractor specs, that’s the scariest part of going into a bid. You’re agreeing when you turn into a bid that you, if you’re a low bidder, you’re going to sign that contract and agree to everything in that thousand pages of specs, so much of it’s boilerplate. You can look at it the other way.

But when something’s hidden in that contract and it gets into a lawsuit over it and you’ve agreed to it because you didn’t see it, that’s terrifying to me. Just how long do you have to do the job before liquidated damages? So many subcontractors are tied into the liquidated damages of whatever the general contractor comes up against.

And then you have to say I have to go to court, decide how much of that applies to me versus someone else’s inability to do their work. It’s it’s amazing. And like you said, it doesn’t take place of good counselor advice at all.

[00:17:21] Wade Carpenter: But it can be a good start. And, if you’re not going to take the advice to go talk to a good construction attorney, then, at least start with that and say, what do I need to know about this? Play devil’s advocate? How can I get hurt by this? And just telling it. And you would be amazed about what it will come up for.

[00:17:41] Stephen Brown: It’ll come up with all sorts of things that you may not have thought of is what you’re saying.

[00:17:46] Wade Carpenter: Right.

[00:17:47] Stephen Brown: And that you needed to know. I get it. That’s great, Wade. This is good information.

[00:17:52] Wade Carpenter: I know there’s a lot of stuff. I wanted to go more into, we’ve used it so much in these, chained little bots that will we can create our automation, that’s what I had started talking about last time, is we have these specialized little bots that will do a certain task, and now we’re putting this in human in the loop, what I referred to as human in the loop, Where we can save a lot of time, but maybe we ask a question and it’ll come back and give us an answer and we have a follow up question, or we can automate some of the research on the company we’re working for or subs or, whatever, that kind of stuff.

And again, we’ve started, as a natural progression of us building all these type systems for contractors for being in their back office, trying to be as efficient as possible. That’s where we started building some of these automations for contractors and making them a lot more efficient and making us more efficient to where they have information a lot better at their fingertips. And, we can pull up some more timely job cost information.

One of the things I was going to do was actually start doing a series just showing some of these automations that we built. And a lot of times they’re not as hard as you might think, or you may just need a little help and realize how much time and effort and cost it could be saving you.

[00:19:12] Stephen Brown: I think that’s a great idea. I think the next logical step to our podcast on automation and AI are some case studies. And also, as always, feedback from our listeners. What are you using AI for?

[00:19:25] Wade Carpenter: Absolutely. If any of our listeners have any thoughts or feedback on today’s episode, we’d love to hear from you. Whether you have questions, insights, or topics you’d like us to dive into next, drop them in the comments below.

Wanted to thank you for tuning into the Contractor Success Forum. For more resources, tips, and the latest updates, visit ContractorSuccessForum. com. Or connect with us on the CarpenterCPA’s YouTube channel or ProfitFirstConstruction. com for more exclusive content. And if you found today’s episode helpful, don’t forget to share, and subscribe.

And join us every week as we explore new ways to help your construction business thrive. We’ll see you next time.