ep: 14 [AI]NTELLIGENT BUILDS: A Conversation About AI at Cerris

Join host Sydney Wickliffe and guests Luke Kotz, Senior Vice President of Technology, and TJ Huettenmueller, Enterprise Applications Manager, for a conversation about a relevant topic impacting our work and our industry: AI. In this episode, we dive into the current landscape of AI, safe and responsible use, and what lies ahead for this tool shaping the way we do our work. 

Check out Cerris' AI Governance Document. 

Explore Microsoft Copilot.

Transcript

SPEAKER_00 0:11

Welcome back to Behind the Build, CERS' internal podcast where we highlight the people, projects, and programs that make our company extraordinary. I'm your host, Sidney Wycliffe. Today we're discussing a topic that's shaping the future of our industry and our company, artificial intelligence, otherwise known as AI. Whether you're already using AI tools in your day-to-day, still figuring out how to make the most of this technology, or haven't explored it much at all, we've got some good information from our company's experts to support our use and decision making when it comes to AI. Joining me today are two key leaders in our AI strategy, Luke Kotz, our senior vice president of technology, and TJ Hutton Miller, our enterprise applications manager. Let's get into it. Luke, I want to start with you. There's a lot of buzz around AI right now, and with that comes excitement, but also some hesitation. We're all at different stages of adopting AI into our work. Some of our employees are probably using it on a weekly to daily basis, and others may have done very little exploration, if any at all. But I think it's really important that we acknowledge that we all bring our own thoughts about AI to the table. So, Luke, from your perspective, what would you say to someone who's maybe more skeptical about using AI?

SPEAKER_01 1:33

First of all, Sydney, thank you for having TJ and I on to talk about this topic of AI. It's obviously very mainstream and across every newscast you see and most commercials and things that you're you see out in the marketplace, you just can't escape AI. And it is a topic that uh is also something that's not going to escape this construction industry. So really excited to have this conversation. Uh, first of all, I would say it is absolutely normal to be on all ends of that spectrum. Uh, from excited to terrified to unsure. That is okay. Um, you're not alone. There's there's really not a one-size-fits-all approach to learning this technology. And every role will counter it a little bit differently. Um, you know, AI is not about replacing people, it's about empowering them. So it's a tool that helps us work smarter, not harder. Um, if we can automate repetitive tasks, we can free up time and focus on things that we do best, like solving problems and building with precision or delivering quality, things that we talk about every day out in the field. Um, and this just isn't a tech trend. I come back to that. It's it's a shift that's reshaping not just our industry, but many industries, all industries really. Um, and in construction, AI is projected to grow at a compound annual rate of almost 30% a year through 2030. That's according to McKinsey. That's a that's a lot, right? That's a big growth number. It just speaks to the fact that this isn't going away. Probably in my career, it's the biggest shift since the internet. Um, and and I truly believe that um with AI and the power of it, um, that it's that transformative. Um, so smarter planning, safety monitoring, accurate estimating, it's all this stuff is is coming out and is already here. It's accelerating, and it is totally normal to not really know where to start.

SPEAKER_00 3:30

Absolutely. Generative AI like ChatGPT or Copilot seems to dominate the bulk of conversations on this topic. But several other use cases for artificial intelligence exist in and around the tools we use every day. TJ, will you talk to us about what AI looks like in our day-to-day? What are some of the tools and systems we're already using that people might not even realize are powered by AI or that have an AI component to them?

SPEAKER_02 3:58

Absolutely. So you are correct. Generative AI, like the chat GPT and the co-pilot, definitely get a lot of the attention. Um, especially around the mainstream. That is what a lot of the topics are about is that generative AI. But we're seeing AI show up in a lot of the tools that we use day to day. So, whether it be in our project management platforms like ProCore or Autodesk Instruction Cloud, those have a lot of AI built into it, both on the front end to help users retrieve data and ask questions about it. But they're also starting to come out with automations and stuff on the back end. So, where if you're doing one thing in one module, it might go ahead and make an update somewhere else. Apps like OpenSpace uh is very AI heavy, so they use a plethora of algorithms to help map the captured images of the floor planes and the good models. Our accounting department uses a software called Cloud CR. It's responsible for reading, validating, and then entering over 120,000 invoices into Keysile each year for your approval. Even something as simple as email of Microsoft Applet that uses AI, it prioritizes importing emails to the top of your inbox, it can help summarize your emails, and you can suggest replies. And then even things that you would use in your personal life, like your maps, Google Maps or Apple Maps. You know, when you plug in your office address or your daily queues or the next service side or job site you're going to, it's very AI heavy to predict the traffic and show you what your best route is. We also have a few more tools coming out in the near future. Um, document Crunch is one that is a contract analysis tool built specifically for construction. It's going to help our users to review and negotiate the new contracts that we have. And it also has a project playbook that not only allows users to chat with the contract and get questions answered that they might have, but also will help it to automatically draft language for any required notices. So generative AI is definitely the fancy and the flashy tool right now, but there's another whole layer of AI that quietly works in the background that's helping to make our operations safer, smarter, and more efficient.

SPEAKER_00 6:15

Yeah, I think that's a really important thing to focus in on is the fact that there are so many other applications for AI that exist around us all the time that we're probably using and not realizing it. And to your point, are making us safer and more efficient in our work. This technology is constantly evolving, and we've seen all the headlines around the push to adopt AI, but there's definitely a balance there. Luke, from your perspective, where would you say that we fall in the adoption curve? Are we more on the leading edge? Do we fall more in the middle of the pack, or are we late adopters? And how does this align with our strategy around AI?

SPEAKER_01 6:57

Yeah, absolutely. I would say as a as a whole and from an overall strategic standpoint within IT, we tend to live in the middle of that adoption curve. Um and that's really where we want to be. We think that's a smart place to be. We're not on the bleeding edge, and we're also not somebody that's falling so far behind. I would say that AI is really no different there. Um we want to kind of be in that middle area, which is pretty common for many companies, and we're still working towards. If you look at things like the Gartner hype cycle, Gartner is an organization that does lots of uh research around all technology, not just AI. The hype cycle is something that they they have termed uh for all technologies, and it's it's this cycle that you'll see when something gets released. And so it'll track sort of these emerging technologies. AI has already passed kind of the what they call the peak of inflated expectations, uh and really is entering into the what they call the trough of disillusionment. So those are big fancy terms, but basically it means we've we had all these expectations about AI, and everybody was all excited, and we're kind of at the top of this cycle, and now we're kind of going down the down the hill into this trough, and and that's where really the hype settles and and real and sustainable use cases start to emerge, right? So you started to hear about everything was AI, and now you're starting to hear more about things that are actually providing value and and coming out of the market. So us being in the middle means we're not jumping in blindly during that sort of hype phase and and when everything's going crazy. And we're also not lagging behind because we're paying attention. So we're learning from those people that are early adopters, and there are companies that certainly did that. And we're watching what works, we're watching what doesn't, um, and and in that process, starting to build our strategy that we think is thoughtful, secure, and aligned with our long-term goals. So it's about timing. Um, that really falls in with all technologies, not just AI, but specifically with AI, I think it's it's vitally important that we that we do time that right so that we don't fall behind, but also don't make bad decisions long-term.

SPEAKER_00 9:14

Yeah, for sure. So we just heard about how we've had this opportunity to really learn from those early adopters. TJ, you've also participated in a local AI roundtable with peers from other companies. What kind of value has that brought to your understanding of AI? And how has our tech team stayed sharp and informed in such a fast moving space?

SPEAKER_02 9:39

You are correct. This is a very fast-moving space. I mean, construction in general has seen unbelievable growth in the past decade just with technology. And now when you throw AI into the mix, things just keep happening faster. They're not slowing down. And like Luke said, AI, just like most new technologies, it presents the unique challenge in understanding what's impactful and what's going to change the industry, and what's just what. And to your point, we find that pure learning is about the best way to understand those differences. Nobody is shy in this industry about sharing feedback, positive or especially negative. Um, so sitting on these committees that are specific to the AEC industry, we get to learn what these other companies are doing. We can benchmark ourselves against them, and it gives us that direct feedback. So if we if someone finds a product or a use case for AI that's changing the way they do things, we'll add it to a list and we'll look into doing that. But on the flip side, if they've tried and they've failed at something, that just tells us that we need to approach it in a different way from how they did, or maybe we just avoid it altogether, and that's a it's a future thing that it's not quite ready for. In terms of our team staying up to date, it's really not as difficult as you would think because we're a bunch of nerds. Um the AI movement is fascinating to us, so both in a professional and a personal capacity. I know that I use AI as much my personal life as I do here at work. Um, and internally in our own team, we use several cutting-edge AI tools for development. So that not only helps us stay up to date and be more efficient, but we can learn as we do that. So things are changing quickly and they're getting more complicated than ever. But I'm I'm very confident in our team's ability to stay up to date.

SPEAKER_00 11:36

Absolutely. And I know that you guys, and we'll get into this a little bit later, but there will be some opportunities for others to also stay up to date right alongside you or to learn from you, which is exciting too. All right, let's shift gears to talk about responsibility. We are a 100% employee-owned organization, which means that we all have a stake in ensuring our collective success. That includes protecting our data, systems, and proprietary information. As part of this intensive work that you've done to set us up for success using AI, you spent a lot of time developing guiding principles that not only anchor us in our strategy, but also guide our use in a safe and responsible way. Luke, can you walk us through those principles and why they're so important to have in place as we move forward?

SPEAKER_01 12:28

Yeah, as with many areas in our business, it's really important to have a good sound strategy. And as we sat down early this year to really start to outline what that would look like, um, we started talking about what is our vision and mission with relation to AI. What are these guiding principles, the things that are going to drive how we make decisions and how we prioritize AI opportunities, um, started to be talked about. What are those responsible AI principles and how do we govern that stuff? And I know we're going to talk about the governance document a little bit more, but in terms of the guiding principles, these are the things that we're going to look to when an AI opportunity comes up and sort of be like the table stakes or the things that that are going to drive yes, we should do that or no, we shouldn't do that. So uh I'll just walk through those kind of one by one and give a high-level overview of them. Um, these aren't just technical, these are really around strategy and they're they're really strategic in nature. So the first one is around aligning with our business strategy. So our our SBUs and our center all have strategic uh plans, and we need to make sure that our these AI opportunities are in alignment with those plans. So this isn't just a side project, it's really embedded in everything that we do or will be. And so, how do we fit AI into that strategy and help with the growth plans and with the things that I've heard Jason and Dustin talk on previous podcasts about? So the second one is around innovation and continuous improvement focus. Is it something that's gonna propel us forward or give us an advantage or be innovative in the space? That we think is important. We want to do that. It's part of our strategic plans to do that. And so we want to make sure that AI obviously is pushing that forward. And most of the time, AI is probably gonna fall in that innovative approach because it's gonna be new to a lot of people. Um, but that's gonna be a focus of ours. Um, the one of the, I would say, big discussion points we had as we put these together was around the third one, which is leverage and applications versus build our own. And what that really means is we're gonna be thoughtful about AI and what we do both internally and how we use them in tools that we have that we're already using. Uh, TJ brought up Procore and ACC and these systems that we're already using that are releasing powerful AI technology into those tools. We believe it's it's prudent of us to take advantage of the AI there and really make use of that AI within those tools before we go build something big on our own. Doesn't mean we won't build on our own, but if we're looking at opportunities, we would err towards is it in applications that we have versus not. And then lastly, around systems, our systems and data are resilient, secure, and responsible. This is something that has got to be uh a table stake in this. We we want to make sure when we're using AI that it's something you can trust, that it's something you can believe in, that's something that's gonna provide you value and is gonna be secure. And you know, with AI, the terrifying part of AI is that you can get to information very quickly with some of this generative stuff. And that can be make things um pretty terrifying when you start layering that over aspects of our data. So we hold a lot of data that's personal data, company data, and things that we wouldn't want to be out there in the world. So we're trying to be very thoughtful about that. These guiding principles are something that I think are are are gonna be talked about as these opportunities come up. And I'm really excited about that, the fact that we were able to land where we landed.

SPEAKER_00 16:20

Absolutely. Your team has worked really hard to ensure those guiding principles and all of our strategies around AI have guardrails to keep us using this technology in safe and responsible ways. Luke, earlier you mentioned the AI governance document. Earlier this summer, this governance document was released. TJ, can you tell us a little bit more about what this document is designed to do and how our people can use it as a resource moving forward?

SPEAKER_02 16:49

Sure. Releasing that AI governance doc was a really important milestone for us. Um we aim to utilize the advanced AI technologies to transform our construction processes, but to do that, we need this government doc so it establishes the guidelines and best practices so we can use the AI responsibly and effectively. We want to make sure that we're ensuring the data quality and the privacy and security throughout the whole AI life cycle, and more or less just make sure that we are using good AI. So, really, the document is a resource. So whether someone's experimenting with a new tool or they want to integrate AI into one of the workflows, it gives them a reference point to make sure that we're staying aligned with our values and also our legal obligations. So, for example, in the document you can find some approved tools like Microsoft Copilot, and it specifically calls out how Copilot is the only approved AI model to use with a company data, but you can still use others like Chat GPT or the Apple Intelligence on your phone for your general use. It's also a living document, so as the technology evolves, so will our policies. You'll see new applications popping up there all the time. For example, we're we're getting close to rolling out an enterprise-wide meeting AI application for meeting notes and summaries. And when that's ready, you'll see communication directly for that, but also it'll live in the governance doc as well. So if someone's looking to use or wondering if the meeting application they're using is approved, that document is out there. It's also to help protect ourselves. Um we're actively putting together a steering committee that the governance doc talks about. That committee will be the team of us who will review and approve or deny new applications or use cases that have been proposed to us. So we're going to make sure that we have a centralized place so that we can know about the uses of FAI across the company. And if we do approve them, again, that'll add it to the document and be made available so others know maybe something's already approved and maybe we already have licenses for something they're trying to do. So, really, whether you know you're an IT or you're an operations or even a field employee out in the field, uh, the resource is there to help support the responsible AI adoption across the company.

SPEAKER_00 19:14

Wonderful. In some recent communications, we've introduced the idea of our digital PPE. Can the two of you talk to us about what it means and what some examples of our digital PPE, including Microsoft Copilot, may be?

SPEAKER_01 19:31

Yeah, so I think about digital PPE, um it's really our way of thinking about safety in the digital world. I I have said uh and been quoted, and I'll say it on this podcast, uh, that risk in the digital world and cyber risk more specifically is one of the top three risks to our organization. Um so we need to be thinking about that just the same way we think about hard hats and safety vests on job sites, uh, and the protective measures that we take out in the field to keep all of our employees safe and going home every day. We need to take those same measures with data and technology and be thinking about it. Obviously, data and technology has been around a long time and is going to become even bigger with AI, no doubt about that. And so, how do we protect ourselves? So we do a lot of things in this area. TJ's going to talk about a couple as it relates to AI, but if you think about other things that we do on the cyber side, it's things like the phishing campaigns that we do, it's the multi-factor authentication that we have, um, it's single sign-on that we're we're doing with applications so that we know who you are and it's easy to log in, but you're not having to, you know, it's those types of sort of everyday IT uh table stakes or protocols that are really fall into the same category. It's all, as I look at it, digital PPE that we've got to have to keep ourselves safe, not just ourselves, but our data, your personal data, our company data, and all everything that lives within our environment. And so it's it's really important. And I just I love that that um parallel to what we're doing in the field every day, which is is putting safety at the forefront. And and I'm thinking about that in the same way with digital. So TJ, I know you wanted to maybe mention a couple others from uh from an AI standpoint.

SPEAKER_02 21:32

Yeah, I think Copilot's the biggest one here. So Copilot's the approved AI model to use with any company data. And the reason for that is because models like Chat GPT are not private. So whether you upload an Excel spreadsheet or even information you enter directly into a chat with Chat GPT, can immediately be found by anyone, any other user over the entire world. So if you upload something like the labor rates or estimating data inside a chat GPT, it immediately goes into the model to train it on and it can be retrievable by anyone else. Copilot, by contrast to that, is private. So because it is licensed through our Microsoft account, as long as you are signed in with your work email, it stays private to us and the info you put in cannot be retrieved by anyone else. So that's a very important distinction that we need to try to get across, and that's that's the reason for all of this. There's also security flaws in some of these AI models. I saw one actually just this morning, the ChatGPT, where if you haven't hooked up to like your Google Doc account, that there is a security flaw that someone can access your Google Docs through Chat GPT, even if you haven't used that to interact with it. So there, as AI keeps moving faster, so do the bad actors. So the security becomes even more important in the ways that we can protect ourselves about it.

SPEAKER_00 23:00

There's a lot we have to do to stay diligent about our use of AI, and I really appreciate you guys and your team for keeping an eye on all of that and providing us with all of the tools that we need to be safe in the digital space, including all of the digital PPE that we're expected to use at this company. All right, so we've touched on the ever-evolving nature of AI. TJ, can you talk to us about what's next for Saris when it comes to AI? What can team members expect in terms of training, tools, and opportunities to get involved?

SPEAKER_02 23:36

Sure. So as I mentioned earlier, we do have a few new tools that we're working on rolling out and implementing across the company. I'm sure there will be communications coming out about those when we're ready. We're also starting to work with the learning and development team to put together some AI training classes with a special focus on prompt training. So prompt training would be how to interact with these generative AI models. Much like humans, there is an art to communication with these models. So while AI is good, it still needs that information from you, so it helps you be the most efficient. That's actually one thing that we learned from the peer groups we sat in. There's one company we talked to who designed and implemented their own chatbot after a couple years, and they were not getting good results with it when it finally got rolled out. And they couldn't figure out why, so they dove in and started to look at the prompts and questions that people were using to interact with it. And they realized that the very vast majority were just one word. So instead of asking questions, people are just typing one word into it and expecting to get stuff back. And it it's AI is good, but it's not that good. So as a lesson learned from that company, we're gonna really be emphasizing this training on the front end of it so that as these tools become more available and rolled out to everyone, people know how to interact and use it to the best of their best of their ability. We also fully recognize that the best AI opportunities and really software and tech opportunities in general are gonna come from the SPUs. So we're we're looking into a way to collect those ideas, almost like a suggestion box. If there is an opportunity that you see across the business, or even just for repetitive tasks that you keep doing, we want to know about it. As these technologies become more widespread, there's probably going to be ways, or there's existing ways already that someone else is using that we can implement for you and to help using AI. So for now, you can send ideas directly to myself or our VISCAHelp at Sarah's.com email address. But we are working on a more permanent way to collect and then track these ideas, and we'll be communicating those when we find out what that is.

SPEAKER_00 25:57

Luke, as opportunities for trainings and new AI applications become available, how do we continue to implement AI in a way that's thoughtful, strategic, and aligned with our values?

SPEAKER_01 26:09

Yeah, well, as I mentioned, AI isn't a trend. It's not going away, it's really a transformation, right? So it's reshaping the way things are done and redefining roles in some cases, um, reimagining ways to solve problems. It's just not going away, right? Uh I did some research. According to McKinsey, again, uh a firm that does a lot of studies around the construction industry. Over 70% of organizations are already using AI in at least one business function. And 40% plan to increase their investment in the next year. Uh in construction, AI is expected to drive over $4 billion in productivity gains by 2030 based on their research. That's a big number, $4 billion, right? Um, and so you know, at Saris, I think we're not just chasing trends. We're really thinking about building the extraordinary and building for the future. Our long-term vision is rooted in responsible innovation, integrity, and quality of craftsmanship when we think about AI. Um, as part of that uh future is built on a strong foundation. Standardization and consistency is super important, right? So the software we select, the cleanliness of the data that we have is critical. Um, these aren't just IT decisions, they're strategic choices that ensure that we use these AI tools effectively. Um, we can scale them, they're aligned with our business goals. So there's that foundational element, I think, that we're working on in doing that, trying to do that in a in a really quality way in order to really be able to use AI into the future. Um our data is clean and the systems are consistent, AI just can do its job better. It can be more valuable. And so that's why we're we're really focused about on that. Um, and that in in increases the accuracy, it reduces the risk and all those things. Um, but as TJ sort of mentioned, we need ideas, we need people that are willing to experiment, we need people to think about things differently, and that really brings in that entrepreneurial spirit that we talk about a lot and is is is really uh key for our organization. Many of you out there are are bringing that every single day when it comes to construction and coming up with new ways. I I see the Living the Values Award nominations and things like that. Like this stuff is happening. We need people thinking about that way as it relates to AI so that we can help figure out what are those things that give us that that competitive advantage into the future. So I'm I'm really excited about it. It definitely has great ties to our values and in every aspect of the way that we're trying to roll this out and listen to you as the business about what you need and what would drive value.

SPEAKER_00 29:03

This has been a wonderful conversation. Luke, TJ, thank you both for joining me today and for the work that you're doing to help us build the future with AI-driven insights and excellence. What this episode really hit on is the fact that AI isn't about replacing what we do. It's about enhancing it. It's about working smarter, staying safe, and continuing to deliver quality craftsmanship that we're known for. AI is a great tool, and we encourage you to get curious with an entrepreneurial spirit. This is an opportunity for each of us to explore new things and work in partnership with our Insights and Integrations team to continuously develop our AI capabilities. If you want to learn more, check out the AI governance document on the Blueprint or reach out to TJ or the Vista Help Desk with questions. Thanks for listening to Behind the Build. We'll catch you next time.