AI in Pre-Construction (How Technology Accelerates Estimating and Frees Humans for Strategy)
Most of a project well done happens in pre-construction. So today I’m talking with Aaron Kivitz from Stack Construction about AI, labor shortages, and how technology is changing pre-construction. Aaron, would you mind doing an introduction, your background, kind of what you’re passionate about?
Aaron’s Background: From Architects to Contractors
Sure. Yeah, so my name is Aaron Kivitz. I work for Stack Construction. I am based in Kansas City, and I’ve been in the construction tech industry for almost 20 years. I kind of started off working with architects, which was a lot of fun, and I’ve kind of switched teams and now I work for contractors. So, we develop takeoff and estimating solutions, and that’s kind of where my passion is right now.
What was fun about working with architects?
You know, architects are artists. And it’s different, right? I got to travel the world, you know, visiting amazing firms like Foster and Partners, you know, seeing models of the Apple campus as they were working them in the office. And, you know, it was just a lot of fun being able to kind of be involved in that creative atmosphere.
The Contractor’s Biggest Need: Technology Adoption
When you’re working for contractors, what is their biggest need and how do you provide that need?
Yeah, that’s a good question. You know, I think we’ve probably all heard by now that construction is the least technology adoptive industry out there, right? I mean, that’s been around for a long time. But I think that’s really true. It’s kind of dramatically true.
I was at ConExpo last week. We had a booth there and I’m still talking to people that are doing takeoffs with a pencil and a ruler and estimating with a calculator. And there’s just so much opportunity to change these people’s lives so that they can spend less time doing things that they don’t really want to do. Just like an architect doesn’t want to put together a project budget, a contractor doesn’t want to spend time estimating.
And to your point, that’s what they should be doing, right? Like an architect should be doing the artistic side of it. And builders should be able to supervise in the field and do what’s important to them so they have capacity.
Construction technology should be doing the things that it should do so we can do the things that we should do, right? AI, like I am a huge fan of AI. Like I use it to help me, not replace me. Yeah, 100%.
Aaron’s Approach to Pre-Construction
What is your approach to pre-construction? And where are you seeing most of the benefits?
Yeah, I mean, I think my world view in pre-con, pre-con is kind of a big word, like it encompasses a lot of things. And my kind of pre-con world is smaller. I’m kind of looking at everything in that estimating lens and how things kind of radiate out from that.
And I wanted to kind of connect this back to what you were just saying a minute ago, like nobody really wants to estimate, but you probably know better than I do how vital an accurate estimate is to the rest of the project. Like if you have messed that up, you’re digging yourself out of a very deep hole.
So, I think this is an area where something like AI, we see is going to allow a contractor to spend less time focused on doing something that they don’t want to do, and more time building things. So, I definitely don’t look at AI as, and it’s not going to replace anyone’s job here. The key is trying to accelerate what it is you’re able to get done, what it is you’re able to do. So that, again, you can get back to doing what you want.
How Stack Uses AI: 12 Million Pages Per Month
Would you mind letting me put you on the spot a little bit with, what are some of the specific things you’re using AI for?
Yeah, so we’ve had an AI-based takeoff tool for almost two years now. So, we’ve been able to do things, pretty basic things like draw walls, draw areas, find windows, find doors, things like that. And we’ve learned a ton from the way that people have actually used those tools. And this year we’re going to see really dramatic advances in what we’re going to be able to do with that capability.
Just to kind of give you an example, Jason, we process over 12 million pages per month through Stack. So we have a ton of data. So we’re going to be able to do things here very shortly, like automatically identify the wall types, automatically identify the materials and the details and the spec for those wall types, and then automatically be able to pick the items and assemblies to go with that wall type.
So that in a few clicks, you’ll be able to go from a set of plans to a pretty much complete takeoff to an estimate with those item and material and labor quantities in it. Like you said earlier, it’s not going to replace one’s job. Like it’s not 100% accurate, right? Like we’re looking at this as a way to accelerate that process.
Rather than you having to get all that work done, you’ll just be able to go in and check and make sure that the computer has done it correctly. Has it missed a wall, right? Has it misidentified a wall type? Was it not able to find the fasteners, for lack of a better example. So you can think of it less as a completed estimate and more as a mostly completed estimate, right? And you’re the 100%.
What Humans Still Do: Strategy, Not Takeoffs
So let me put it in my own words. We’re still going to have conversations of where we source materials. We’re still going to have conversations about does the general contractor take this or that scope on as self-perform? We’re still going to have conversations about which trade partner we should select. We’re still going to have conversations about what is the right assembly? What would be fit for prefabrication?
But we shouldn’t, if I’m hearing you right, have to worry about, do we have the right takeoffs? Do we understand the specs? Do we understand what is actually included in an estimate? Like these basic things that AI right now can help us with bid leveling, with estimating, with quantity takeoffs, with all of these things that allow humans now to go do the strategy portion better to get to the overall budget so we can even build the job. Do I get any of that right?
Yeah, I think that was right. And you can do all those things in Stack today by hand, right? You can draw a wall or you can assign an assembly to it, and you can calculate those material quantities. The difference that AI will make is, instead of you having to do it, it’ll do it for you.
It’s not going to understand how difficult it is to build that one. It’s not going to understand if the architect shouldn’t have put the wall there, or whether or not it is the right type of construction. It’s not going to replace the experience that a seasoned estimator has that can figure that stuff out by looking at a set of drawings and understanding those facts.
Narrowing Feedback Loops: From Weeks to Hours
But what I’m hearing from you is that the experienced estimator will still do their job, but it’s going to narrow the feedback loops, right? The time that it takes. Like, typically in design, you’ll get a new drawing set, and then it’s weeks to figure out what the impact is. And back in the day, it would be weeks to figure out what the schedule impact is.
And let’s just even forget about constructability reviews. The team doesn’t even have time to deal with that. So now, if I’m hearing it right, AI is going to help us know, this is the new quantity, this is what impacts the budget. With things like Takt planning and other Lean scheduling systems, we’re going to be able to see within hours what is the scheduling impact. And now we actually have the capacity to go do the builder type things that we’ve always actually been paid to do and need to be doing.
That’s a good way to put it. And I think it’s really, I kind of look at AI just as a tool to do exactly what you said. It’s trying to, it’s going to mean I can spend less time doing the low value work and more time doing the things that really require me to think through a problem.
Addressing Labor Shortages: Multiply Skills and Influence
And you mentioned that there is a shortage of trained project delivery team members in the industry, as well as there’s a shortage of trained, skilled labor in the industry. And if somebody’s like, “Oh, you’re going to replace my job with robots,” well, no. Teams complain all the time about being overwhelmed with admin. Can you talk to me more about that? How does your approach help fill in those gaps?
Yeah, it’s a consistent message that we hear from our customers that they just cannot find estimators to hire. They want to bid more work. They want to expand. They want to grow, but they can’t find estimators that can help them bid those jobs.
And really what they’re sort of left with is training people to do it internally. And that process, I mean, that’s obviously vitally important that these contractors out there are passing on those skills to the next generation. But that sort of on-the-job training is really how you become a seasoned estimator, right?
And I think the key there is, instead of replacing the estimator, you’re going to spend less time clicking around a polygon to draw an area on a plan just to measure it. And more time helping somebody else learn how to then do that constructability review, to understand how the schedule would be impacted by the change. That’s really, I think, how AI and tools like this can help with that labor shortage problem.
Basically, we’re increasing capacity. And the folks that use a service like Stack or are using AI are going to multiply, meaning multiply their skills and their influence faster and do better in their career.
Yeah. And I think one of the reasons why there’s a labor shortage in this area of the industry is also just, it’s filled with these dated tools and point solutions that are disconnected. And you fire up something like Bluebeam and it’s a 20-year-old desktop application that has no numbers in it. It’s just there. Why would you want to do that? And then they’re cutting and pasting numbers into an Excel spreadsheet somewhere. And these estimators have these Excel spreadsheets that they’ve worked in for 25 years.
The Problem with Disconnected Workflows
Whenever I’m on a video, I’m like, I love Excel with the possible exception of getting a formula wrong in an estimate, which we’ve all been there. It’s like, “Whoa, whoa, whoa, whoa, where’d that 700 grand go?” Well, I had the wrong formula in the Excel sheet.
And I think the problem with those tools is because of that disconnection between the takeoff, the quantification and the Excel estimate. If there is a change, you have to start all over or you have to pull this thread back from your estimate to where on the drawing that thing was, figure out what’s changed. That sort of disconnected workflow is just a gigantic waste of time. 100%, 100%.
Here’s what AI in pre-construction enables:
- Stack processes 12 million pages per month – AI-based takeoff tool for almost two years. Draw walls, draw areas, find windows, find doors. This year: dramatic advances. Automatically identify wall types, materials, details, specs. Automatically pick items and assemblies. Few clicks: set of plans to complete takeoff to estimate with item, material, labor quantities.
- Mostly completed estimate, humans check accuracy: Not 100% accurate. Accelerates process. Rather than you doing all work, you check: Has it missed a wall? Misidentified wall type? Not able to find fasteners? Think less as completed estimate, more as mostly completed estimate. You’re the 100%.
- Humans still do strategy: source materials, trade selection, assemblies, prefab: AI won’t understand how difficult to build. Won’t understand if architect shouldn’t have put wall there. Won’t understand right type of construction. Won’t replace experienced estimator who figures that stuff out by looking at drawings. Humans do strategy, AI does takeoffs.
- Narrowing feedback loops from weeks to hours: Typically, new drawing set, then weeks to figure out impact. Now: AI helps us know new quantity, what impacts budget. With Takt planning and Lean scheduling, see within hours what scheduling impact is. Now have capacity to go do builder type things we’ve always been paid to do.
- Addressing labor shortage: multiply skills and influence: Customers cannot find estimators to hire. Want to bid more work, expand, grow. Instead of replacing estimator, spend less time clicking polygons to draw areas. More time helping someone else learn constructability review, understand schedule impact. Folks that use Stack multiply their skills and influence faster.
If your project needs superintendent coaching, project support, or leadership development, Elevate Construction can help your field teams stabilize, schedule, and flow.
Aaron’s Closing Statement: Don’t Be Afraid of AI
Anything else you want to communicate?
Yeah, I mean, I think the key takeaway is that Stack, we’re trying to solve these problems. We’re trying to give people more time in their day to do the work that they love. And I think that these tools like AI shouldn’t really be something that you’re scared of. It’s the people that are taking advantage of it that are going to win. Not the people that are afraid to use it or don’t want to use it because they’re afraid of it.
Learn how to use it, learn how to apply it. I think tools like Stack can give you really actually practical usable ways of doing that. So that’s really what excites me is really helping people get back to the work that they want to do.
That’s a powerful passion statement. Like that’s really important in our industry. Well, yeah, I mean, that’s why we show up to work, right? I like to solve problems. Contractors like to build stuff. Architects like to design stuff. That’s why we’re doing what we do. Where can people get ahold of you?
Yeah, Stack, you can find us at stackct.com. We’re fully cloud-based, so you can just go there, create an account, get started. Our AI tools, like I said, they’re going to be dramatic advances this year in what we’re able to do. And we want more people to be able to jump on that train. As we say at Elevate, AI accelerates pre-construction: Stack processes 12M pages monthly, automates takeoffs and estimates, increases capacity. Tools free humans for strategy work.
On we go.
Frequently Asked Questions
How does Stack use AI for takeoffs and estimates?
Stack processes over 12 million pages per month. AI automatically identifies wall types, materials, details, specs, then picks items and assemblies. Few clicks: set of plans to complete takeoff to estimate with quantities. Not 100% accurate, you check. Mostly completed estimate.
What do humans still do with AI-powered estimating?
Strategy: where to source materials, does GC self-perform this scope, which trade partner to select, what’s the right assembly, what fits prefabrication. AI won’t understand how difficult to build or if architect shouldn’t have put wall there. Experienced estimator still needed.
How does AI address the labor shortage in estimating?
Customers cannot find estimators to hire. Instead of replacing estimator, spend less time clicking polygons to draw areas. More time helping someone else learn constructability review, understand schedule impact. Multiply skills and influence faster.
How does AI narrow feedback loops in pre-construction?
Typically, new drawing set, then weeks to figure out impact. Now: AI helps us know new quantity, what impacts budget. With Takt planning and Lean scheduling, see within hours what scheduling impact is. Have capacity to do builder work.
Why shouldn’t people be afraid of AI?
It’s the people taking advantage of it that are going to win. Not the people afraid to use it. Learn how to use it, learn how to apply it. Tools like Stack give practical usable ways. Helps people get back to work they want to do.
If you want to learn more we have:
-Takt Virtual Training: (Click here)
-Check out our Youtube channel for more info: (Click here)
-Listen to the Elevate Construction podcast: (Click here)
-Check out our training programs and certifications: (Click here)
-The Takt Book: (Click here)
Discover Jason’s Expertise:
Meet Jason Schroeder, the driving force behind Elevate Construction IST. As the company’s owner and principal consultant, he’s dedicated to taking construction to new heights. With a wealth of industry experience, he’s crafted the Field Engineer Boot Camp and Superintendent Boot Camp – intensive training programs engineered to cultivate top-tier leaders capable of steering their teams towards success. Jason’s vision? To expand his training initiatives across the nation, empowering construction firms to soar to unprecedented levels of excellence.
On we go


