
Building a Data-Driven Future in Construction with Annette Cooper of Graham
Annette Cooper
Gut instinct alone isn’t enough in construction — data is now a game-changer. Annette Cooper, Director of Data and Analytics at Graham, sits down with us to explore how data is transforming the construction industry. She explains how Graham leverages analytics to manage risk, optimize efficiency and improve project execution. Annette emphasizes that success comes from solving real business problems rather than simply generating reports.
I'd rather read the transcript of this conversation please!
Key Takeaways:
(03:17) Graham is expanding rapidly through acquisitions, creating new data challenges.
(04:05) Risk management is a priority in an industry with long project timelines and tight margins.
(07:26) Many project managers are eager for data insights but lack access to the right tools.
(09:05) A standardized data strategy is essential for integrating multiple systems post-acquisition.
(12:45) Project estimations require a balance of historical data and real-world unpredictability.
(17:22) Effective data teams focus on solving problems, not just collecting data.
(20:01) Data should drive decisions, not just generate reports that frustrate users.
(25:42) Predictive models can highlight risks early, but human expertise remains crucial.
(31:55) The future of AI in construction lies in automating insights, not replacing workers.
(35:10) AI can streamline building code compliance by identifying design issues before construction begins.
Resources Mentioned:
New Zealand Ministry of Business, Innovation & Employment
Anette: [00:00:00] My line in this end is, if this isn't helping push our business forward, if this isn't helping us solve problems that we have, then why would we do it? That's a waste of resource.
Anthony: Welcome back to Data Masters. Today I'm speaking with Annette Cooper, a leader who has spent the last two decades turning data into strategic assets across a variety of industries from financial services to government. And now in the construction sector, she's built and led data and analytics teams that drive real business impact.
As the director of data and [00:01:00] analytics at Graham, she's at the forefront of leveraging data to enhance decision making in large scale, municipal and infrastructure projects. Graham is a construction firm. That's been shaping skylines and communities for over 90 years, taking on complex projects across building industrial.
And infrastructure sectors in an industry where precision, efficiency and risk management are critical. Data plays a crucial role in optimizing their operations, improving safety and ensuring projects are delivered on time and on budget. But as we'll discuss, it's also an industry where people have traditionally relied on their gut or perhaps a phone call to a foreman to understand where a project sits.
Her career has taken her across New Zealand and Canada, leading research and analytics teams at organizations like ATB Financial and the New Zealand Defense Force and the Ministry of Business, Innovation [00:02:00] and Employment. She brings a unique perspective on how data can transform decision making at scale.
I'm excited to hear her insights on the intersection of data. And also the construction industry where challenges and opportunities in applying analytics to infrastructure projects are bound. So welcome to the show and at
Anette: Thanks for having me, Anthony. I'm excited to be here.
Anthony: awesome. So I think a good place to start. we've, I think it'd be fair to say I have never had anyone from the construction industry on data masters. and I'm not sure. Most people really understand how this industry works, and what makes it unique and different. Probably many people think of the construction industry like they might think of hiring a contractor to renovate a bathroom or something like that.
Your business is very different than that. And, and alsoyou come to it. with a, rich history and experience across, as I mentioned, a variety of different places. So maybe just share a little bit, who Graham [00:03:00] is and how you, ended up there.
Anette: Yeah, so, I'm not surprised to hear that I'm the first from a construction company. we are traditionally perhaps, further back on the wave when it comes Um, but I also think that's what makes it exciting. So Graham Construction is one of the largest construction companies in Canada. And we're in the top 50, although that may have changed through a recent acquisition.
We may be further up that list in the U. S. too. Primarily on the West Coast in California and Washington and all across Canada. we do. major construction, so skyscrapers,bridges, major roading, paving projects. We have a specialized water division that does water treatment plants and we don't fix people's kitchens.
We don't help people, decide on paint choices and things like that. So we're big civil and big engineering projects. That's what Graham does All across North America. what [00:04:00] interested me and what got me here was the fact that I am the first one of me at Graham. So Graham was in a place where as a construction company that started to realize that data were becoming more and more important, particularly for risk management and construction projects. But the Capability that they had in house was small and in pockets, primarily focused on financial reporting, and that it was time to take a wider view and it was also time to get more strategic. Right. Like how do we use data to help us manage risk across construction projects? Construction is a low margin business with very long timeframes.
So the kinds of stuff we do, if you think about how annoyed you've been by a roadworks project in the middle of your city, that's really only half of it. You know, we've got a year before that Of us thinking about how we're going to upset your morning commute. And then the actual process of us doing that. So these are long [00:05:00] projects low margin. So any issues in execution can very quickly eat into any profit that we have. So how do you help use data to. manage those projects and manage that risk. me personally, yeah, I'm from New Zealand. That's where this, accent comes from. And I have had a very unintentionally long career in and around construction type stuff. So when I worked at the ministry of Business innovation and employment, which is a government, central government agency in New Zealand. I worked in building, construction policy and housing policy and building systems.
So all of the government regulation that goes around large scale building. And then when I moved to. The defense force, which is the New Zealand combined militaries, we have, they're the largest, government owner of infrastructure in New Zealand, right? Huge amounts of land for bases, camps, they run an airport, they run a Navy, you know, so we've got one [00:06:00] of everything and a lot of. time and effort and energy trying to use data to help them manage physical assets for life cycle, like asset life cycle management. So I did this accidentally, but I think the thing that got me interested in Graham was the kind of green fields and the freshness and the excitement to, I think, Show that construction is different to how people traditionally think of it.
A bunch of people driving around in trucks with hammers in the back that were actually quite a sophisticated, sector, and that there's a lot of opportunity here for data folks and technology folks.
Anthony: Yeah. So I think that's, what I see is a common theme, both in your, experience in New Zealand and then with Graham is the, these are, Very large dollar value projects, but they're very project oriented. So presumably, unlike projects in Boston, which have a beginning and a middle, but [00:07:00] never an end. I'm sure your projects have a beginning, a middle and an end, but you know, so they're project based.
so once you've built the skyscraper, presumably you're done and you don't build it again. There are large volumes of money at risk and at play. and then I think it's also fair, but I wouldn't mind your perspective here. the there's this interesting combination of, very on the ground, hands on work.
People are literally putting things up, pouring concrete, doing very, you know, big things. and yet, you they need to be data driven. is that a fair characterization of the business project oriented big dollars and maybe less sophistication around data, especially on the ground?
Anette: Yeah, I would say at a business level, that's true. I think we have a lot of individuals on our sites who we probably don't give enough credit to in terms of their sophistication and their frustration, particularly our project managers, their frustration that they can't get their hands on the data they know they need, data that would help [00:08:00] them. So you get a lot of requests for, can you help us pull together information on this project because we know it's there, we just can't get our hands on it. So I think. there is quite high data literacy. It's the technology maybe that's not quite in the place it needs to be able to service those needs. I would say it's more that. Yeah.
Anthony: Yeah. So, that's better said. So really, it's a strong desire for data, but the kinds of processes we're involved in are maybe difficult to measure and hard to get a handle on, you know, the exact metrics associated with them. But before we go there, and I do want to talk about that. Maybe, help ground us.
we know you'd mentioned Graham just finished an acquisition. I think one thing we talked about before was that Graham has really grown through acquisition. that's always a challenge when you're looking at a data infrastructure and a data landscape. and maybe share a little bit about when you joined Graham.
What made it such an interesting [00:09:00] greenfield operation, pun intended, since many of your projects start with greenfields. So
Anette: so when I joined Graham about three and a bit years ago, we were, probably maybe. Two thirds of the size that we are now. So we've grown that much, in terms of revenue in three years.what really interested me. And continues to be the challenges, the acquisitions, right? So how do we set up a data platform an enterprise data platform that can be agnostic to or systems?
So we're in SA, the Graham Classic, as we call it, is an SAP shop. So we have SAP, but then every time we buy another company, they have something different or nothing, which is the other interesting challenge. But how do we architect and think about not just what would help us now as a company, but in five years time, if we can project ourselves down that road and we continue to grow by acquisition as we have, then what do we need to be thinking about in that first [00:10:00] sort of six to 12 months to set ourselves up to be successful in five years and not to be constantly making the today decision, but making that. That three year down the road decision so that we can be pleased with ourselves when we get there that we made the right one, rather than kicking ourselves that, you know, we knew this was going to be a problem and we ignored it. So we went about building a data platform that was agnostic to any of our large ERPs and one that was specifically designed for data.
And the other thing that. Even though primarily now we're doing numerical information, we wanted to hold space for unstructured as well, right? We want to be able to get into unstructured. We want to be thinking about building plans, drawings for things, and then being able to compare them to as they are built and things like that. That all becomes really important when we're managing schedule performance, which is. That's the thing that all major construction are driven by and anything that involves [00:11:00] the renovation of your home is also driven by the schedule, right? Like, I want my kitchen in five weeks like you said it, the city of Calgary wants its major ring road in three years like we said we were going to get it. that's been the part that has been really interesting and exciting And I think how we're knowing Graham will grow and how we're responding to make those plans now, rather than waiting for them to happen at us.
Anthony: yeah, I think this question of growing through M& A and thinking through a data strategy that's consistent with that is a theme we see quite a bit across a number of different industries and in any M& A deal idea of building synergies across the two organizations is always a primary driver of why the two organizations merged And then that's typically driven by some data, like looking at overlap across customers between projects, materials and your in your example, I'm sure. [00:12:00] And actually, which sort of brings an interesting idea. We wouldn't may not necessarily think about the construction industry as being seeped in data, but it sounds like this is, you know, in a way, your problem has been too much data.
There's too many different forms of data, too many different E. R. P. Systems. There's a lot of data here. that a fair way of framing it? Like it's almost like an embarrassment of riches, like so much data. And now the question is how to bring it together. Is that fair?
Anette: yeah, I would say though that old adage that 20 percent of our data answers 80 percent of our questions is really fair. And then the frustrating part is always why do we not have these two fields, which are actually critical to how we're going to understand something. And we have all this other stuff that we're trying to chop away. Graham has been particularly lucky in that 20 odd years ago, we built a project management tool.and so the tool it's called toolbox, is very old now. It's looking at it in Carter CD [00:13:00] ROM when you sort of put one of those in and it chugs to life. but it worked incredibly well for what Graham needed it to do and it. Collected all sorts of data. So we have a huge historical project database and we are transitioning off toolbox now because much like in Carter, it has come to the end of its life and onto a different more sass based project management tool. And now the challenge is how do we take all of that historical data and match it together with. The data on the new platform so that we can have, continuous view of projects, but also how do we understand projects outside of traditional construction. So we also own what's called a services business. So we do maintenance and turnaround where they like shut down an oil refinery, and then we'll go in and do all of the maintenance in a six week boost. That's very different to a traditional construction project where we've got a fixed bid or whatever. This is very much time and [00:14:00] materials. We send a crew in, they do the work, we come out again. So now we're trying to understand, well, how do you match those two things together to get an understanding of performance across two really different kinds of Contract structures, but yeah, we've got a ton of data and I think like any other organization in any other sector, it's about getting to the questions that we really need to ask of it and not doing dashboards for dashboards sake. Not making pie charts and throwing them up everywhere. How do we really get to the root of what's a problem here that we can solve? What's a way that we could take this data and inform operations on the ground, shovels in the ground, about where they're at versus where another project's at, or where they're at compared to historical Projects that look like these. So that's the kind of stuff we're trying to get to.and I don't know, I'm sure I'm speaking to other data nerds, so it's a safe, it's a safe space, but that [00:15:00] notion of. we have to get to the problems, like we have to get to the business problems and then how we solve them. We don't want to be stuck in the business of just pumping out reports that no one's reading or, spending a lot of time building dashboards that look amazing. Are valueless to the organization? I think that's been, I wouldn't say a big challenge, but that's part of the challenge. It's like we build the infrastructure, we get the data, but can we get our key business leaders into the room to have conversations about real problems? And then can we also keep their interest in the work that we're doing long enough to come on that journey with us, because it's really a single conversation.
it's often three, four, six months worth of conversations to get to what are we building and how do we keep refining it? And how do we keep making sure you have something that's actively helping you on your projects?
Anthony: Yeah. And so this [00:16:00] idea of a partnership with business side of the business, is always an important part of making any data project successful, maybe share a little bit about, how you've tackled that problem. And, again. relating back to the industry. they might not have the right language to talk about what they want how to, translate, from the challenges they [00:17:00] face in the field, to how you build a data strategy that maps to that.
Anette: there's a bunch of ways I can answer that. I'm real people person, which I think makes me somewhat unusual, in this kind of a role. I'm not saying that's not true of others, but I think it's one of the things that's made me, helped my success. and so we just take a people first approach.
So I have people in my team who are very much out in the business facing folk, working with people to help understand what their problems are. And I'm also not afraid of a no and not afraid of a, well, how about we think about it differently? What if we don't think about it as, you know, I'm frustrated by X, but we think about. What would help me to move that along? And is there data that we can collect? I mean, I often have a lot of conversations that go somewhere along the line of, well, that would be wonderful if we collected that. We could, you know, we could really help. Do you want to [00:18:00] collect that? Because then we can look at building a mechanism for its collection, but at the moment we don't.
So is there anything else that we could,proxy or we could work on together or something like that?
Anthony: do folks on your team. Literally go to construction projects and like, how embedded do they get,
Anette: We primarily come in more around the district managery level. So that would be someone who has, senior PMs and then PMs reporting to them. That's probably the person that, that kind of level that we would talk to. But we have also like in our people space, we've been working on reporting of indigenous people working at Graham.
So we have Graham employees and then we have a huge amount of craft workforce who don't, who are contracted into Graham, right? But trying to get an understanding of both indigenous people working on our sites and then indigenous businesses that we're subcontracting with. At that point, we are working with Very much that sort of project management level and people on site. I mean, it would be pretty funny to [00:19:00] see a bunch of my data engineers out with their hard hats on. I would like that. And I think I might actually do that now you've suggested it, but no, we're mostly at that sort of district managery VP
Anthony: that is the extreme. but getting back to this, you asked, I think, an important question, which is, if you want to know the answer to that question, then here's the data we need to collect to answer that question. I think we often think about these data questions from the perspective of.
Well, here's the data I have here. The questions I can answer. You know, it's like we think data out. You're thinking about it from the perspective of the question informing what data we collect. maybe share a little bit more about practice how that's worked for you guys.
Anette: My preference whenever we're talking about creating data products is to say, what do we need to know? And my background, you know, way a million back is as a researcher, right? So, We're going out to find out the answer to a question, because we need to [00:20:00] know it for some reason. When I was working in government, we needed to know it to inform how policy was going to be built. But there's no data, or there's a little bit of data, but it's not telling us the whole story, and it's not enough to make a decision on. that's just my training, and then that's the approach I've bought into this because I think when you start from the data, often people get really frustrated.
It was like, well, this is what we have. I'm like, okay, but is it actually helping you? I mean, you've spent six months building a Power BI report that you're annoyed that you have to manually refresh. So is this delighting you? Are you able to make like changes to your business? Well, no, because what we really want to know is this other thing. Okay. Then let's find out the other thing. Right? Let's not do that thing. Let's find out the other thing. And we can either go looking for that data, maybe we do have it somewhere, or we can design a mechanism for the capture of that information, and if it's important to you, you will capture it, right? And then we can do our part on the other end [00:21:00] and build you the Power BI report that actually answers your questions and actually helps you make a decision. Rather than the one that you sort of look at and get frustrated with. So I think it's just, in a world where you do have a lot of data, you need, you know, like I could be building all sorts of things with the stuff that we have. My line in this end is, if this isn't helping push our business forward, if this isn't helping us solve problems that we have, then why would we do it?
That's a waste of resource.
Anthony: love this idea that you come at it from the perspective of a researcher. think, as data teams, we often come To these questions from the perspective of a technologist. here's what I'm able to build you with the extreme version of that. And perspective you're bringing is,what's a, it's like a research project.
and in that context, you may not be building, you know, a permanent analysis that is run every single time. But you may be more like, let's answer this question that [00:22:00] may or may not end up as a permanent fixture in the project planning process or whatever the question is that you are answering.
Maybe thethe, almost the advice or the thing that, you know, a listener could take away is to refactor data work away from trying to create permanent artifacts that avoid interaction with people. Like, I think our dream is like, let's create a dashboard that people can go to, and then we never have to talk to people again, would that be great? As opposed to like, let's go and kick off a research project with. By necessity means we're really talking to people like really a lot.
Anette: Yeah. I mean, curiosity, right? When I was a kid, I pulled my mother's hairdryer apart and then tried to make a disco ball turn with the motor. So I think it's part of who I am is just this curiosity about, well, how do things work? Well, why is it like that? Right. And I agree with you in some levels, I would love a single dashboard that did everything we ever needed. And I never needed to talk to anyone again, but it's just not [00:23:00] reality. And every part of our business, like all other businesses is a little different. whether it's the differences between how we run a vertical building project versus how, our infrastructure team works differently. Buildings do a lot of subcontracting and infrastructure do a lot of self perform. So there's differences, right? So you can't just assume, well, it's like that far over there. So it must be the same over here. You have to get curious about the differences and things. End. I also think there's no size of data team that's ever going to be able to do everything that an organization wants. So if you're not able to be curious about are we really spending this precious resource that we have on the things that really matter, then I personally wouldn't find that very fulfilling just to be doing busy work. Like we want to know that we're leaving the place in a better state than we found it.
Right. [00:24:00] The campsite rule. Um, So that's really driving for me. Are we really making a difference? Are we really helping where we can help?
Anthony: Right. And again, to link it back to what you'd said before, that means connecting this back to business value, the questions the business needs to answer. And those are more like research projects and less like, infrastructure projects or the poor choice of words, but more or less like, building dashboards for dashboard sake or building reports for report sake.
Anette: Yeah, absolutely. one of the things that we're really focused on at the moment is project execution, and because we have across all of our business units has two kind of main phases, right? There's the bid and estimate phase, and that's before we've won any work, we're trying to figure out How big is the project, you know, and it's at the top, at the point where we get to an estimate, we have to be pretty sure we know exactly what's involved because if the estimate is accepted, then that's what we're working to, right? And then we have the build [00:25:00] phase, which is in the execution. So the two kind of major research areas we're looking at is one, how good are we at the estimating? And I would say. The estimators themselves have a lot of information about, you know, how much for the per square foot of what kind of blah, blah, blah, but it's the part at the end where we come in and thumb it up and down a little bit, well, we want to get about this much margin.
So we need and we need contingency because. A five year project, of course, something's going to go wrong. Again, back to, you renovate your house, you pull off a piece of drywall and oh. Right? So it's that on a multi million dollar scale. so it's that part that's less scientific, I would say. looking at that and then looking at when we execute. How well have we, firstly, how well have we estimated? And then secondly, things we can see relatively early on, so sort of 25, 30 percent of the way into construction, that [00:26:00] would give us a sense of whether or not the project was actually on track or whether it was already starting to go. Off track, because the quicker we can get things back on track, the quicker we can protect, that low margin, so it's been there has been a very interesting and it's also the part where we're starting to look at some predictive modeling, like can we build out a predictive model once you get to 30%, what does the Machine predict that you're going to finish it where you estimated or not? that is a really interesting project that we're working on and benefiting from all of that historical data that we have on project execution. but all of it in the mind frame, I think of, yeah, of this idea of this is a research project, right? We have to tease out all of the factors that. Impact,project execution. So we could have two of the exact same projects and one goes poorly and one goes well, because one of them just happened to have a, you know, I'm in Calgary, a Calgary [00:27:00] winter and. They were delayed four weeks because of a snowstorm, and then the whole thing's off, right? So there's all of these interesting, intricate parts of this that we're having to really tease out and think about. And then how does the data then take shape to tell the stories about what has impacted project execution historically? And then how can we use that to project, knowing that we can't actually project a snowstorm as much as we might like to believe we can.
Anthony: Yeah. I want to connect this back to the sense of people because, and your sense of like estimators are having a little bit of a wiggle room in their estimates. but I also imagine what I may be asked more as a question. Do you find some level of resistance to say, predictive model, uh, whether a project will complete on time or not?
people react?yeah. Negatively to that. Like there's always humans are, relentlessly hopeful. Like, yes, we're behind and yes, the model predicts that, [00:28:00] but we'll make it up. You know, like, or how do you find the, especially specifically in the, with regard to that endeavor, mapping it to the human element to do, you know, how do you get people to trust an estimate or a prediction, a predictive model?
Anette: I mean, people are inherently sceptical of my role generally, right, because I'm not a construction person and I think that's where the good people skills help, right, because I don't come in with the intention of saying that I am. I'm here to bring a specific set of skills and my team is here to bring a very specific set of skills to a problem that also requires the subject matter experts to be sitting in the At the table and saying, Oh, actually, we're going to remove that project from the sample because we couldn't do anything about that, there's no storm. So that is actually an outlier, right? So like identifying those kinds of, patents, but like all companies, we have resistors and, I would say the executive leadership [00:29:00] and,that VP the possibility predictive modeling. I think that as you get further,out into operations, I wouldn't say that they're necessarily always resistant, but I think that they would have a healthy dose of skepticism. to throw at that kind of work, because yeah, it can show things that otherwise have been able to be hidden. we have at the moment, a project oversight. you know, process, but of course it's all lagging indicators of, you know, projects in flight. And we also have projects that have cost our company a lot of money. So I think the executive are a little off the mind that I don't really care. We need to find these things before they happen. So whatever we need to do to find The issue projects, then we need to find them. And I would also say like, [00:30:00] sometimes what gets flagged is just projects that aren't putting their data in the system. So having that being brought up is okay. So there's actually nothing wrong with the project. Like the project is executing fine, but they're not putting their information in the system. Well, that's actually still a problem. Like that's now a behavioral problem, right? We need system adherence so that we can do this work. So. I'm personally delighted when we pick up those kinds of projects because that means one, it's working, like there's a system adherence issue. but also it's helping to get that message out about this is what this data is being used for. Right. And I think sometimes in an organization our size, people don't understand always, you know, like They're putting things in and getting frustrated with the 10 places they're being asked to put the same bit of information in. And so being able to spend the time educating about what, when you put that there, this is what we're using it for. So when you don't do it, this is why you're getting flagged. Up is read on our [00:31:00] project controls dashboard because you haven't put your stuff in. So that stuff excites me too, right?
That being able to show what information is being used for being able to show that the data are being looked at, understood, used in all these different ways. I love it.
Anthony: So, when I cast your eye to the future for a moment. So everybody's talking about Last I checked, we will not be replacing people who pour concrete and weld steel with AI agents that I'm aware of. You can correct me if you think I've misstated the facts. but generally, like, where are you? do you see as the innovative future projects?
you mentioned, unstructured data, but where are you in investing, what do you see as coming? How do you think about AI playing into that? does it have a role? I'm curious on your perspective on that.
Anette: I think that, AI is a very broad term and [00:32:00] for us at the moment, there would, there's definitely productivity gains that could be made, in processes and things like that. Some of the more exciting use, like in the data sides, I'll just talk about the data side. So no, we're unlikely to replace, people pouring concrete.
Although we can do things like. Automated,old drilling, you know, and stuff like that, right, where the, whatever the giant machine just sort of goes along as the pattern is laid out for it. So we do have some of that kind of stuff, in the data world. It's really, can we utilize RAG type, AI solutions to help get business people answering ad hoc questions that they might have of data that's in our platform, right?
Like those questions that get asked of, well, how many people worked here in 2020, whatever, or, what was the profit in buildings? Canada and quarter three of that kind of stuff, right? Which is really time consuming for people to go and look for. So [00:33:00] being able to set up chat genie type things that can take some of that, work away from our, early out financial planning people, we're looking at stuff like that.
We're looking at other things similar to that, to, put across the top of, you know, What we refer to as our grand management system, which is all of our processes and procedures across the business, some of which are corporate service focused, and some of which are actually project execution focused. So people can just pipe in, you know, where am I supposed to be like, gate six or whatever internal stuff. and then BIM, building information modeling, which is a sort of very construction y, uh, digital twin type, work. That's where I think, there could be some really neat use cases for AI in terms of, well, this is what it was supposed to look like and this is what's been handed over. What are the differences?
Like, can you compare Builts, I was [00:34:00] actually chatting to a guy about this not that long ago and,saying, having an AI that could go across the plans for a, you know, we're talking about say a 50 story tower. That's got five under the ground as well in going through and finding, having an AI go through and find problems in the design.
So things like this hallway is technically wide enough on the specification, but when we put a handrail is we must for accessibility. Now the hallway is actually too small. And we don't pick that up until we're halfway through. Realizing that we've poured the concrete and these are not going to be wide enough. So there's things like that, right? We're using an AI to go through the thousands and thousands of pages and to take things like the building codes and all of that and go through and do match and have a look at stuff like that. I mean, I think that would be game changing for us, isn't it? As an industry, not just as an organization, because once you have built something, it is very costly to unbuild it.
Anthony: [00:35:00] That is, I'm sure, something that all listeners, from their personal experience with construction will remember that, much easier to change in the plan than once you built it.
Anette: Yeah, and when we're talking particularly about public use spaces and, places where we have really rigid building code around. Accessibility, as we should, fire, egress, like all of these things have their own kind of
thick, this is true in New Zealand and I spent time on it, something that could take all of that and then look at your plan and go, yes or no, here's, you know, here's what's going on. I think those are the kinds of things that, I would love to see AI working on, as opposed to, you know, the 900 offers I get a week for the same rag, we can fix your everything, which you can't.
Anthony: feels like a somewhat solved problem. Like those are not overly complicated, use cases.
Anette: Yeah.
Anthony: Annette, thank you so much, for your time. I really appreciate, a unique [00:36:00] perspective. And as I said, like, I don't think we've had anyone from this industry, on the podcast, but. I do think there's a lot of commonality across what you shared things like both, long running projects, people who, are used to, trusting their gut and then now looking to look at data.
I think the idea of running these as, I liked your idea of research projects, that's a really nice way for people to think about connecting the work they do. Back to real problems, because, presumably a research problem is researching for some reason, that's a problem. and then, you know, I do agree with you in terms of the context of AI and thinking about, really looking at the places where people spend a lot of time and energy, to your point about building codes, as opposed to these really simple use cases we see, people talking about today.
But thank you, very much for the time.
Anette: Thank you for having me. This was really fun and hopefully helpful. [00:37:00]