#318 Olek Paraska: How AI Is Fixing the Biggest Bottleneck in Construction

Construction is one of the least digitized industries in the world, and not because it resists technology. It resists bad technology. In this episode of Eye on AI, Craig Smith sits down with Olek Paraska, CTO of Togal AI, to break down why construction productivity has barely improved in 50 years and why pre-construction is the real bottleneck holding the industry back. Olek explains how most estimating and takeoff work is still done manually, why automating this phase can unlock massive efficiency gains, and how AI works best in construction when it acts as a perception and reasoning layer rather than a replacement for human judgment. The conversation explores computer vision, agentic AI, human-in-the-loop systems, and why respecting real-world constraints is essential for AI to deliver real ROI. It also looks ahead to a future where floor plans, materials, costs, and constructability can be reasoned about together, long before construction begins. This episode is a deep dive into how AI can finally move construction forward by solving the right problems, in the right order. Stay Updated: Craig Smith on X: https://x.com/craigssEye on A. I. on X: https://x.com/EyeOn_AI(00:00) Why Construction Is Desperate for Better AI (01:06) Olek's Path From Software to Construction (02:17) Why Construction Productivity Has Stalled for Decades (04:33) The Pre-Construction Bottleneck No One Talks About (06:17) How Takeoffs Are Still Done Manually (09:15) Why Construction Rejects Bad Technology (11:18) How Togal Found the Right Problem to Solve (12:14) From Computer Vision to Reasoning AI (17:44) What Agentic AI Looks Like in Pre-Construction (20:59) Turning Floor Plans Into Materials and Costs (28:18) The Real ROI of AI for Contractors (47:11) The Long-Term Vision for AI in Construction

Speaker names marked ? are inferred from speaking time and the show’s known hosts — they are not voice-verified.

• 0:00Craig S. Smith?

  Togal itself has like tripled in revenue, right? In the last, I don't know, last few years.

• 0:08Speaker 01

  We've tripled in revenue three years in a row.

• 0:11Craig S. Smith?

  Yeah, annually tripling in revenue. To be fair, is that because you're a startup? You're growing from a small base? Is that indicative of how hungry the industry is for a solution like this?

• 0:23Speaker 01

  I would say the latter. I think the industry is hungry for a solution like this. The industry is hungry for someone to listen.

• 0:33Craig S. Smith?

  So usually I start by having the guest introduce themselves. You know, explain your background so far as it's relevant.

• 0:46Craig S. Smith?

  You know, I imagine you've been doing this for a long time.

• 0:54Speaker 01

  Yeah.

• 0:54Craig S. Smith?

  You've built AI systems across multiple industries. So can you give your background and then I'll, I'll start asking some questions.

• 1:07Speaker 01

  My name is Oleksandr, Oleksandr Barasova, also known as Olek Barasova. I'm a CTO in Togal AI right now. And at Togal AI, we built AI systems for construction industry. But before that, I was building AI systems for end tech, advertising, where we deployed AI to millions of users before it was as widely distributed. And before that, I was in a streaming media industry where we built a digital rights management system. Completely different angle.

• 1:41Craig S. Smith?

  You were in, in building digital, what kind of systems?

• 1:44Speaker 01

  Digital rights management. So it's just.

• 1:47Craig S. Smith?

  Oh.

• 1:47Speaker 01

  Ensure that.

• 1:48Craig S. Smith?

  That's interesting.

• 1:49Speaker 01

  Only that people have access to the content.

• 1:54Craig S. Smith?

  Yeah, yeah. So at Togal, can you talk about?

• 2:02Craig S. Smith?

  What's going on in the construction and the housing industry?

• 2:06Speaker 01

  Yeah.

• 2:07Craig S. Smith?

  What, yeah, how Togal addresses it, how AI can apply to it? Uh, yeah.

• 2:17Speaker 01

  Sure, uh, so it, I stumbled into Togal a bit of a by accident, but also because I was looking for something like this. Um, back in two thousand twenty, I think I quit my job back then. I decided you need to start a startup, but I didn't know what the startup should be about. But I did know that I'm, I'm looking for a, a area where AI would make any meaningful impact. Um, the reason like construction is interesting, I think, is because in the last fifty years, uh, the improvement in construction from any sort of innovation has been zero. In some cases, it has been negative. So we have all of this progress in, in the world, in the world of bits.

• 3:00Speaker 01

  Uh, but in the world of atoms, it's going the different direction. And for me, as a software person, that was, um, interesting and challenging. I didn't understand why is that happening? And right now it's all coming to the head, but you know, it's, it's something that has been ruined for decades, obviously. But housing crisis is a symptom of that. There's obviously problems with financing, problems with, uh, all sorts of things with zoning. Uh, but it also is true that we're just not building fast enough. And that's what Togal is trying to tackle. Uh, it, it seems trivial, but before construction actually starts, there's years of, um, negotiations, who, who does what?

• 3:45Speaker 01

  How do we actually build? Because you only get to build once. And that, those years of negotiations and like, uh, understanding and planning can be significantly, uh, decreased into like maybe weeks, months by AI. Because that is all a very decomposable, um, uh, process where AI can deliver a lot of value, right? And this is what Togal is focusing on. We are focusing on pre-construction. So after the architects have finished designing all of the fancy buildings that they want to build. But before construction starts, this process is ripe for innovation. And this is where Togal is. On this spot.

• 4:26Craig S. Smith?

  Yeah, that's interesting. So what what kinds of activities pre-construction?

• 4:33Speaker 01

  Yeah, for example, when you have a collection of floor plans, right, this is what typically is being sent from architects. What happens, and I couldn't believe it when I first heard about this, that every construction person, every subcontractor and general contractor would measure things on those floor plans manually. So everybody needs to understand, oh, how big is this room? Because I need to put tiles in this room. I need to put carpets into this room. And both the tiles person and the carpet person will just go and manually measure the room on a floor plan. And when I say measure, literally, the way it looks like is they go and open a file in some program and they trace around the room. And that's how they measure.

• 5:16Speaker 01

  It's not the most efficient process, but for every building, even the smallest building, it happens hundreds of times because there's hundreds of subcontractors. And that seemed like an inefficient process. So this is the process as we are automating. Instead of you having to draw all of these polygons and lines and and count, you know, how many doors you have, how many toilets you have. We give it to you for free because that's where AI shines. And then you can focus on higher value tasks, like you don't want to be tracing things. That's what kids love to do. You want to design, like, understand how to build this pattern. So stuff like this. Uh, from tracing around, from being a perception layer for uh construct uh construction.

• 5:59Speaker 01

  But also there's other higher value tasks that now are possible with intelligence. Uh, like, understanding what is the scope here, like, maybe doing some reasoning. It's uh it's ripe for many, many places. But we started from uh with something called takeoff. You are measuring all of the things on the floor plan.

• 6:20Craig S. Smith?

  And and so the the user um gets the the drawings from the architect and uploads them to Togal's

• 6:32Speaker 01

  Exactly.

• 6:33Craig S. Smith?

  Interface, I guess, and and it will immediately count all of the doors of and give their dimensions, count all of the rooms.

• 6:42Speaker 01

  Yeah.

• 6:43Craig S. Smith?

  Give their dimensions.

• 6:44Speaker 01

  Yeah. Yeah, we will, we will immediately you will upload uh the floorplans to us, but you typically people upload all of the construction documents to us because there's floorplans, but there's accompanying spec books and there is like hundreds and hundreds of pages of text description of what is it that we want to build. So yes, we are a perception layer. We look at the construction floorplan. We have proprietary computer vision models that extract the measurements from those floorplans. You have all of the lengths of the walls, all of the uh areas, all of the counts for the doors and and windows and and stuff like that. That is given to you immediately. But if you want, you can also do this manual as you are doing this before before, because

• 7:29Speaker 01

  we are sure that, you know, AI is AI and you might want to count something differently. And so not only do we give you AI tools, we also give you the regular tools that you have as a manual take of estimation. And that hybrid of how much manual, how much AI is always, you know, a challenge. Because AI can do much more. Maybe you don't want to do a lot of things manual. But our philosophy is always to put you in charge. So you are always uh in the know, you are always in control. One of our slogans is you are the captain of the ship. So we can give the tasks to your mini minions in total, but you are the captain and you also are responsible for your uh decisions.

• 8:15Speaker 01

  For us, if you misestimate, if it's like a huge building, if you misestimate, the value that error can be millions. And so that's uh in a way similar to still driving cards industries. The the value of error can be very, very significant. So that's why there is a balance that you have to strike into how much agent.

• 8:39Craig S. Smith?

  Yeah.

• 8:39Speaker 01

  If.

• 8:41Craig S. Smith?

  Yeah, uh in the estimating, I mean, I'm also surprised uh the way you describe it. So they get the architectural drawings, the the contractor or the subcontractor, and they literally measure things by hand, I mean, you know, on a computer screen, but uh.

• 9:04Craig S. Smith?

  By by tracing the outline of of the room or something uh or or counting the number of doors they need to is all of that manual today?

• 9:15Speaker 01

  Yes, and I don't want to, you know, diminish the uh value of of this because in some sense, even being able to do this on a computer screen is already an achievement in some cases because there's also still people could do it in pen and paper. They would print it out, they would take out the ruler, and that's just the process that they follow. And I'm I'm not saying this in any sort of derogatory form. My point is, construction is one of those industries that touches the physical world.

• 9:46Craig S. Smith?

  Yeah.

• 9:46Speaker 01

  Coming from like an advertising or streaming media, it's very, very different and you can feel it in the kinds of people who are there. Uh for the lack of a better word, and I'm sorry, I'm using it in a technical philosophical term, uh they are very uh averse to any sort of bullshit. And so because they are dealing with physical realities, they would not buy something that is uh not going to be helping them. And people say that construction is resistant to technology, but I found that construction is resistant to bad technology. And that is like a majority of this, and this is a big, big problem. And it's very hard to satisfy a need of a person who's working in on night and day with their hands.

• 10:32Speaker 01

  And technology is the last thing that they think about. And that's the the kind of difference in construction that you have to work with.

• 10:39Craig S. Smith?

  Yeah, and so this bottleneck that you're describing, project estimating, you said it's called a takeoff?

• 10:46Speaker 01

  Takeoff, right.

• 10:47Craig S. Smith?

  And why why is it called a takeoff?

• 10:50Speaker 01

  Because taken off quantities from the floor plan.

• 10:54Craig S. Smith?

  I see, okay. And construction productivity has stagnated or even declined for decades. And that is a major opportunity where automation can deliver.

• 11:10Craig S. Smith?

  Real results is was Togol already focused on the construction industry? Is that something that that you saw?

• 11:18Speaker 01

  The Togol started as a spinner from the largest general construction company in South.

• 11:25Craig S. Smith?

  I see.

• 11:26Speaker 01

  And so it comes from a deep knowledge in construction. And the way it started, the founder, Patrick Murphy, he's now a CEO of Togo. Was looking into wake up his construction company workers, his family construction company works. And then he identified this as a one of the biggest inefficiencies. Because it doesn't seem like it's it makes sense to keep doing this in two thousand twenty or something. And there was definitely a problem to be solved. And AI was brought in as one of the solutions. It wasn't like, hey, we have AI, let's apply it somewhere. It was completely the opposite, yeah.

• 12:03Craig S. Smith?

  And and you quickly saw how it could be? How computer vision could be harnessed for this?

• 12:12Speaker 01

  Yeah.

• 12:12Craig S. Smith?

  Task? Yeah.

• 12:14Speaker 01

  It started as computer vision. So computer vision is in a way a perception layer. So we teach computers to see and understand to see the floor plans. But now with large language models and generative AI globally, we can connect that perception layer to the reasoning layer that large language models provide. And that is like increasing in the amount of capabilities significantly. Because now that the agents can see what's going to floor plan, they can already understand it because they have been trained on construction knowledge in some shape or form. You can take this much further. And so it's a lot of back office tasks that are very automatable because you can connect perception and reasoning.

• 12:59Speaker 01

  And that's what we are betting on. So it's not just the vision now, it's like the whole package.

• 13:05Craig S. Smith?

  Yeah, yeah. So you moved beyond construction estimating.

• 13:11Speaker 01

  We.

• 13:11Craig S. Smith?

  Into other back office.

• 13:14Speaker 01

  Yeah, we moved from perception of for construction estimating, we move in, we're moving to pre-construction phase more broadly. Because that phase itself, pre-construction industry itself is a pretty broad industry. So we're staying in pre-construction. We don't want to go in any other areas. But there is no, for example, system of record for pre-construction. And that's what we are solving for.

• 13:40Craig S. Smith?

  Yeah, and you started with with all of the floor plans by what Patrick Murphy's company was Coastal Construction, is that right?

• 13:49Speaker 01

  Correct.

• 13:50Craig S. Smith?

  And so you started with all of their plans. How do you train a model on this? Do you create synthetic data or you buy architectural plans from off from public sources? How do you get the training data?

• 14:07Speaker 01

  That is a great question, because that was the one of the chicken and egg problems that we had to solve early on. So we had a collection of floor plans, but they were just that, they were just floor plans. What we needed for to train custom deep learning models, we needed to annotate them. And so we had to find people who would annotate construction floor plans. And it seems trivial, but it's not, and not anyone can annotate a construction plan. It's it's so nuanced and sometimes I wouldn't know what to annotate these things. And so definitely, it's not like somebody on mechanical territory. We had to hire professional architects to annotate things for us, and that's a little bit of complexity. But in the end, we have annotated thousands of floor plans.

• 14:50Speaker 01

  We were able to train our first deep learning models for wall detection and for classification, object detection. And then we were able to kickstart the engine and now we have more floor plans and then we can annotate them again. But in this process, we also have produced a way to generate synthetic data. Because these floor plans are also coming from specific software. And there's only so many users of that software. So we have the process to synthesize any amount of floor plans and automatically get them annotated. Just frankly, from our experience, synthetic data is always worse than real, old, messy data from construction people.

• 15:34Speaker 01

  At least our machine learning models will always be able to tell what is synthetic, what is not, and that's something good thing.

• 15:42Craig S. Smith?

  Yeah, yeah. And this model, is it an open source model that you're fine tuning? Or did you build a model from scratch?

• 15:51Speaker 01

  We build the models from scratch. It's transform, there's multiple models, but some of them are transformer based. So in a way, it's inspired by a transformer architecture, but it's completely trained from scratch by us. Some of them are some object detection models, like U L V six, U L V seven, I think. And so that we took those models and then we tuned them on construction floor plans. But for the main segmentation, we have trained from scratch.

• 16:22Craig S. Smith?

  Yeah. And so is that a smaller model? I was just talking to a company, Cohere, do you know them?

• 16:29Speaker 01

  They're. They're, I'm actually have been very involved in them, yes.

• 16:34Craig S. Smith?

  Yeah. And so they build small models that people can, that people can deploy on site. Is that important for what you guys do? Or are you a cloud based model that they hit for inference via API or something?

• 16:53Speaker 01

  First of all, it's small model in today's terms, it fits in P H P U. But it's a pretty big model in terms of like how deep learning is concerned. But we do not deploy it on premises. We, they hit it, people will use it through an API or through our front end service, which is like web application that we provide. There's many reasons why deploying those models on premise is not a good idea, but not the least of them is in construction, people not want to mess with complexity. They don't want to combat complexity where it's not needed. So why go through the pains of hosting the model if you can just use it through an API?

• 17:36Craig S. Smith?

  Yeah, and you're moving into agentic AI. How does that work in in in construction?

• 17:44Speaker 01

  Uh, specific in preconstruction, there is preconstruction is almost all back office. So it's all operating with some sort of documents and moving documents around. For example, one of the cases is you as a subcontractor for another drywall. You receive a set of documents, but it's not clear for you. Am I supposed to do this or that? And so like, it's a very typical process that you receive a document and then you need to send an RFI for. What what is it like? We want to request more information, but to generate it, RFI, it's like it's a whole process. You need to generate it because you need to log it for future purposes.

• 18:25Speaker 01

  Because later, if you didn't ask for it, or you didn't clarify, you didn't have the logs, you are responsible. That whole process, it's like a, you generate an RFI from a text. Very routine, there's nothing that should not be automatable here. Stuff like that is very susceptible to automation, using agents. Another example is, again, you receive a document, but you just don't understand. What's the scope for you? What are you bidding on? And there is like five hundred pages of documentation. Some of them is relevant for you, some of them you just don't care about. You're sitting in a truck somewhere in the field. You just want to say, what am I looking at?

• 19:07Speaker 01

  And so just parsing all that information, summarizing it for you is a perfect fit for an agent. But we come through and what our core strength is, you can do all of these things using large language models. What we bring in is we add a perception layer. Because now these large language models can understand the floor plan. And so from there, you can just say, here is one version of this floor plan. Here is another version of the floor plan. What are the differences? But what are they meaningfully different? And compare them analytically and just tell me like, what do I need to ask? What am I not even thinking about? Big part of reconstruction is analyzing construction. Sure, architects might have thought about this, but you are a construction person.

• 19:52Speaker 01

  You know what they don't know. Large language models have that baked in, but you need to tap into it in a specific way. And that's another kind of agents, that agent that can shine in.

• 20:05Craig S. Smith?

  Yeah, and you mentioned sitting in a truck. Is this a field centric solution? Or is it back? Are you sitting at the construction company headquarters figuring this stuff out?

• 20:18Speaker 01

  It's typically would sit in a construction company headquarters. It's a back office. One of complaints that our salespeople have is that when people join from their trucks, it's very hard for them to sell to them because they don't, they have million things going on. That happens, but it happens only occasionally. You would typically do your job for pre-construction in the back office.

• 20:40Craig S. Smith?

  I see, yeah. And these agents you're talking about, they can handle things like identifying materials or

• 20:50Speaker 01

  Yeah.

• 20:50Craig S. Smith?

  managing work, close a ton of sleep. Can you talk, just work, walk us through a use case that would demonstrate some of that?

• 20:59Speaker 01

  Okay, so for example, you have a floor plan. You are a drywall contractor, we thought a subcontractor. We talked about this, for example. What Togol gives you out of the box is just a line for where the wall should be. And you know that drywall probably should be a drywall. But the next obvious step for you as a subcontractor, you need to understand through this line for the wall, there's a small tag connected to it. That small tag says something like M4, M5. And you as a subcontractor know that this means this is a specific type of a wall. And I have to go and look through all the other pages, like 200, 2000 pages, and find where is that tag defined to understand what is this wall?

• 21:44Speaker 01

  And you find it in like a page 275. This wall is specific kind of a drywall. It has this many corner bees, this many studs, and it has to have two sheets of drywall and stuff like that. Now that you know this, the next step for you is, okay, I need to connect what I have discovered here to this quantity that I have here. And that's a non-trivial task right now, because again, you have to do this gymnastics. And from there, you need to understand, well, when you are estimating how much would it cost for you to build this? It's not just the quantity, you need to understand what are the materials that will go into this.

• 22:25Speaker 01

  So you need to build all of the assembly for that wall and count every stud, count every sheet of drywall. And it becomes like an optimization problem that is again susceptible to computer thinking and cognition. But it needs to be framed that way. And if you can go from this process from the very beginning, we had nothing, to the end where you have a bill of materials that you might need. This is the ultimate holy grail, and that's a collection of agents that we are building.

• 22:56Craig S. Smith?

  And so the vision is that a contractor can take the architectural drawings, give them to Togal's system, and then query Togal through natural language of how, what's the total cost of this project? In today's market? And it would go through and figure out how much steel girders you need, how much glass, float glass you need, whatever. And it'll spit out a number, a cost number. And then you, can you then say, do the agents, can they order the materials?

• 23:41Craig S. Smith?

  Can they say, you're gonna need X square feet of drywall? Let me order that for you from a supplier, or how granular or active does the agent get?

• 23:53Speaker 01

  Yeah. Do you not have plans to build agents who order for now? There is different plans that will probably be better suited for that. Perhaps a marketplace where people can give bid on the project if they know what they need. It would be better suited. We don't try to go there because that's a bit of a different phase of construction. Again, you try to stay in pre-construction. On the other hand, if you ask about vision, the vision is that where you can map from a geometry of an architect to build up materials in the real world, in the physical world. That is a very powerful thing to have because when you have that,

• 24:35Speaker 01

  you can quickly change the plan and understand how it influences the value of the building. And that's a lot of, that's a lot of things that can be solved by value engineering like that. So for example, for housing crisis, okay, you cannot afford this much. If we remove these things, this is how much it would cost. Ultimate vision for Togo will be like you have a prompt and that prompt will generate a plan for you and then plan will generate a bill of materials. So sure, we might be able to generate those plans, but we still want to have humans involved for like other reasons. So the real value is no matter what the plan is, we can give you the materials.

• 25:19Speaker 01

  And then if we generate plans, if architects design plans, it becomes less relevant.

• 25:25Craig S. Smith?

  Yeah, yeah. And then the contractor or the subcontractor, who, as you said, knows, has a wealth of knowledge in what works or what doesn't work or what's easier or more difficult. Can, can, can they then ask the system, is there a substitute for this material which is not available right now? Or is it that kind of thing where you can? Have a conversation with the system about the building.

• 26:00Speaker 01

  Exactly, yeah. So now that the system knows what is it talking about? In a way, it has an internal representation of BIM, like building information modeling, without you explicitly having to build BIM, because that's an expensive process on its own. So it builds this BIM internally and then you can query it. Some of the questions people would ask is if I want to build this building, but make it more green, give me some recommendations. What would make this building more green? Is it maybe a different kind of concrete or like older concrete from here? Or maybe insulated for, I don't know, for other reasons. So you can then reason on top of this because you have all this information already there and the agent is already connected to it.

• 26:45Speaker 01

  What kind of questions you can ask is up to you. Typically, we see people ask questions about constructability again, because that's important. If an architect forgot to have a drain for a shower, it's nothing for architect, but it's on you because you're responsible. Then you have to tear down the walls and reconnect it. Like, what am I missing is typically what people want to know? Because that's something you cannot even predict. What am I not asking? Other questions, like I said, what is the scope for me? Give me some recommendations. How can I decrease the costs? If I want to decrease the cost, how to make it more luxury?

• 27:25Speaker 01

  Those are the kind of things that you can get into once you have this internal BIM representation.

• 27:32Craig S. Smith?

  Yeah. Does it? I mean, it certainly, I would guess, increases productivity. But does it, I mean, beyond the time savings, does it lead to real ROI on a project? Because you've been able to avoid problems, as you said, see whether or not there are drains in the showers or you've been able to adopt cheaper building materials because this model makes it easy to.

• 28:07Craig S. Smith?

  Test what would work and what wouldn't work. Where's the value for the contractor?

• 28:13Speaker 01

  For the concrete, the immediate value.

• 28:15Craig S. Smith?

  The construction, the construction company.

• 28:17Speaker 01

  Yeah. Oh, for both, for both of them, the immediate value is you would have had to spend a day coloring polygons. You don't have to do it now, so you just save the time. That's immediate value. The second value is other than just having all this fancy AI. It's, it's probably not as important to talk about this in a podcast like this, but you are bringing these floor plans on the web, which most of these construction companies do not have. And now you can collaborate between your, between different estimators. And so that collaboration process is already increasing the efficiency because you don't have to send those files in PDFs, in emails, one to another.

• 29:01Speaker 01

  That itself is a huge value gain. From AI, you just press a button and then you have the understanding of what other quantity is there. It's also a huge value added. The chat, this connected to the application is something is being developed right now, so I cannot give you concrete values for our eye. We know that people use it for quickly, quickly decrease in time for understanding the scope, answering the questions, they just ask a question, what is this thing? And then they have a response immediately. So that's where we are right now. And although there's a lot of stuff to build still.

• 29:46Craig S. Smith?

  Yeah, but again, on is, are there some metrics on how much companies can save or how much inefficiency the system can save?

• 29:57Speaker 01

  So from going from a day to an hour of estimating time. So that's typically say 90% efficiency saving on estimating.

• 30:08Craig S. Smith?

  Yeah, why is construction productivity lower now than it was in the past?

• 30:14Speaker 01

  I can give you theories, of course, and I don't think I'm well qualified for this, just coming from software, that was also a question I had. My understanding is that construction, actually a lot like software, is a lot more about people, systems and less about technology. And in my experience, software is also a lot more about people, systems. But construction, people, systems combined with physical realities makes it super hard to adopt any new challenges because the value of entry is too super, super high. You have to understand construction extremely well to be able to deliver value at least a little bit.

• 31:00Speaker 01

  And frankly, that's also what we're feeling in total, the company. We have people who come from construction, and this is the first software that they are working, software company that they are working in. We have people coming from software, and this is the first construction company that they are working in. And there are these just not understandings of each other inside the company. And we're actively working on trying to figure out how do we talk to each other? Because this is, the worlds are so different. And the way we solve this within the company is also how we solve this for the industry. But in my mind, there is such a big gap between the technology and construction

• 31:41Speaker 01

  that it's, it's a gap that very few companies can jump through. Why is construction not improving in efficiency? I think, again, because it's not about technology as much, it has not been about technology as much, it was more about people, system. And I believe why, right now, we are in a point where we can really change things. Because AI is the only thing that can quickly unravel all of the backlog of innovation, it hasn't been there. And it really democratizes what is possible, but without the gatekeeping of software people. And this is why I'm excited to work on a project like this.

• 32:26Craig S. Smith?

  Yeah, yeah, it's fascinating. It's an industry that has not, certainly it's changed with technology, but it doesn't change quickly. And you're working on massive buildings, I would guess.

• 32:39Speaker 01

  Yes. Primarily, primarily we work on commercial buildings. So that's like big buildings typically it's hotels, hospitals and stuff like that. There's some construction companies in one of the customers, among the customers of Togo, who are building like family, one, two family homes, but typically it's commercial.

• 33:04Craig S. Smith?

  Yeah. And this idea of starting with the estimators, which is fascinating. I had no idea that there was such a role. You would think that would have been estimated, automated a long time ago. But is, I would guess that the architects themselves are using AI in, in, in generating plans. So isn't a lot of that information already available from the architect?

• 33:36Speaker 01

  Yes and no. Architects are using, let's say, digital tools to design these floor plans. But they use these tools only as means for them to accomplish the tasks that they have. They do not think about or can think about things that construction people need to think about. Architects would design a beautiful wall and they would specify, I mean, this is probably a drywall. But exactly how that drywall is built, only a drywall subcontractor can tell you, not even a general contractor, because there's so much nuance when you think about the physical reality of that specific case, that you have to think through every

• 34:21Speaker 01

  detail. And that's why there is like this distinction between different roles in the construction. Architects think big picture, general contractors think, how do you construct this? And then every little trade has their own nuances. You even try to take up a window, there is so many nuances in a window, what can be different, what can be done differently. And that's why it needs very deep specialization. And part of the reason why I think Construction tech has not been the most prominent until now is because in my mind, there was a little bit, there is still a bit of an arrogance from like a Silicon Valley. But we can solve this, this is easy.

• 35:06Speaker 01

  It's extremely hard to, if you think about all of the nuances of the real world. And it's much easier to solve a problem when you are operating a bunch of bits and that's it. But when real world hits you, it's okay. This is, this is much more complicated than you thought.

• 35:25Craig S. Smith?

  Yeah, and the Togal itself has tripled in, in revenue, right? In the last, I don't know, last few years, last three years?

• 35:38Speaker 01

  Yeah, we tripled in revenue three years in a row.

• 35:42Craig S. Smith?

  Yeah.

• 35:44Craig S. Smith?

  Oh, annually, tripling in revenue annually?

• 35:47Speaker 01

  Yes.

• 35:48Craig S. Smith?

  Wow. Is that, to be fair, is that because you're a startup, you're growing from a small base? Or is that, is that indicative of how hungry the industry is for a solution like this?

• 36:01Speaker 01

  I would say the latter. I think the industry is hungry for a solution like this. The industry is hungry for someone to listen. And I would like to think that the Togal is trying to get the best of both worlds. And so we hope that it will continue at that rate.

• 36:21Craig S. Smith?

  Yeah, you're, you guys are a modern tech team or company in another sense that you're a distributed team, right? How do you manage to work across time zones? And are the, is the customer base distributed as well?

• 36:38Speaker 01

  Customer base is primarily United States. We do have. This amount of customers in Europe, Australia, Middle East, but primarily United States. We're starting the global expansion. Communication and remote first culture, I think, are coming from the software side of the company. We are trying to hire the best people for the job and then it doesn't matter really for us where they are based. And so that has been a recipe for us to build a highly efficient team. Doesn't have to be the biggest team, but the team is cohesive. We try to prioritize all of the things that the remote first company would have. We have asynchronous communications.

• 37:22Speaker 01

  We live in Slack and we use, heavily rely on some, on lots of documentation. We do all the things that remote first company would have. We do offsites so that everybody still needs to meet each other. Because if you meet a person, it's very different than you're meeting the square on your screen. But that itself, I would also say, is non-trivial, especially after having tripled three years in a row. The size of the company is growing, so with every few months, it's a set of different challenges. And at some point, the amount of people require new processes. You have to constantly innovate, both in the industry of construction, but also in the way we build this company.

• 38:06Speaker 01

  Because again, two different industries are connected here. So how do we bring both of these industries into the multiple world isn't, isn't obvious. That's frankly what we're thinking about day to day.

• 38:20Craig S. Smith?

  Yeah, and your team, you grew up in Ukraine.

• 38:24Speaker 01

  Yeah.

• 38:25Craig S. Smith?

  But is it a, do you have a lot of Ukrainians in the, on the team or?

• 38:30Speaker 01

  Yeah, we do, I think. A ratio of Ukrainians in total is disproportionately a big, but into a month of Ukrainians in the world. It is, it happened that the war in Ukraine started while we were building, so that expedited some hiring processes. Some of our developers got mobilized into the army, so there was like such a set of challenges if we had to figure out how to work with. But I think that only made us stronger in a way that for a decent amount of people, it's this or the frontline.

• 39:11Speaker 01

  So it's like the agency, the urgency is even bigger for us because you have to deliver.

• 39:18Craig S. Smith?

  Yeah, yeah. And there's a lot of talk of a housing crisis, not only in the U S, but around the world. What do you attribute that to?

• 39:27Speaker 01

  I think it's a, from my perspective, a structural problem that has been brewing for a generation. And now we are just witnessing the policies from the past. I do think that's probably the bulk of it, but I think the other part is, like I said, there is not enough building going on for various reasons. And so if we can only expedite the, how fast we are building, that would solve things at least a little bit.

• 40:07Speaker 01

  And sure, you solve things a little bit and then maybe somebody else solves things from a different angle and it would be much better. I think it's a very complex topic, so I try to focus on what can be done from this side of the equation. And I think we can definitely build faster if we decrease some inefficiency. There's also inefficiencies, for example, in permits. So once you have everything done, you have to sometimes wait a year to just get a permit. There is no reason why there shouldn't be an agent somewhere and gives you a permit in the same day. It's just, it's not there. It's a lot of people's systems that need to change, but people's systems take time to change.

• 40:53Craig S. Smith?

  Yeah, yeah. And this is your, your customer bases in the US. What are there some signature buildings that you can talk about that you guys have worked on? Or you're not supposed to be talking about customers?

• 41:10Speaker 01

  I can't. Some customers we can talk about. I know we have built, we have been involved in building most of the high rises in Miami, for example. That's, there's been like some of the best and fanciest skyscrapers in Miami have been utilized in total. There's been many more and very high-profile buildings that, that you know, high-profile companies that I'm not sure if we can disclose. Uh, just for managing expectations, we have not worked on nuclear power plants and stuff like that.

• 41:50Speaker 01

  Because that's a set of different challenges. Uh, you would need to get FEDRUM certified and have a whole different levels of security. But we are working with a lot of top, top general contractors.

• 42:05Craig S. Smith?

  Yeah, and that's interesting Miami, because I would guess building, a building in Miami is going to be very different than building one in Boston.

• 42:14Speaker 01

  I know.

• 42:14Craig S. Smith?

  And tho- those differences presumably you run across edge case that are not in the training data. Uh, what do you do in that case?

• 42:25Speaker 01

  I think you're right. Building a building in Florida, for example, there's hurricanes and all of that stuff. There are different requirements for the building, there are different codes you have to follow. What is often the case is that even though you know that you're building in this location, you have understanding what kind of codes are there. And so then that's what you need to check against. So if it's Miami, then you have to check against this collection of codes. If it's somewhere in New York, it's different. What are the edge cases, for example? Again, going back to our example of drywall, drywall can be very different in, again, Miami or somewhere north. Somewhere they build with timber, somewhere they build with other kinds of rails.

• 43:11Speaker 01

  And we try to accommodate and build flexibility in our user experience where you can, for example, build a drywall assembly. To your needs and we try to stay out of the way from how you do these things. It's even, it gets even more complicated if you go outside the United States. There's nuances in different countries that you cannot even think about right now. For example, in Germany, you have to show how you arrive to do area calculation, but literally show I multiply this by this and that's how I get my area. Otherwise, it's considered illegal. So like little, little stuff like this, we try to build this into the user interface.

• 43:54Speaker 01

  You can have it if you need to, but otherwise just try to stay away from this. We're not being opinionated. We're giving you a collection of agents, collection of minions that you can use. And then it's you are the captain of the ship.

• 44:09Craig S. Smith?

  Yeah. And then is there a lot of variation? You mentioned codes, is that built into the Total System? Depending on your location, would it surface those concerns or those constraints?

• 44:25Speaker 01

  Total actually has a sister company called CodeCompliant, which uses the same engine, same AI, but it's focusing on code compliance. And that company is working and it has built in all of the codes. So we might say, I'm building this good in here. Please run a battery of tests so that I can understand what is not compliant here. And it's stuff like, what is the shortest travel path from every room? And I know according to the code, it should not be more than this. And if it's more than this, give me a recommendation, what should be changed? Stuff like this. So they handle all of the code compliance stuff and we try to enable them in that.

• 45:10Craig S. Smith?

  Yeah, yeah, that's so where is this going? You're focused on on construction. Are you guys going to look beyond construction? Or is are there areas within construction that you're moving toward? That I mean, beyond estimating and that sort of thing. Yeah, what's on the roadmap?

• 45:34Speaker 01

  Yeah, I think I would like to say that we're moving past construction. But realistically, I think even just the preconstruction space is way too big to even dream to solve. It's such a tangle of all of the different problems and issues and opportunities. That I think our goal is to become a system of record of preconstruction. And that means that instead of you sending emails to all of the different subcontractors, you just go into total and then that's where you power things. But we definitely think that a lot of this stuff is being done now. Might be done with agents or with help of agents.

• 46:18Speaker 01

  And so we want to give every estimator a lot of a lot more agents. That can they can become more efficient. They can do estimates better. They can bid to more jobs. That should drive the costs of construction lower and speed of construction. To make a speedy construction higher. But the big vision is, like I said, someday you will be able to say, I want a building in this place in Florida on the beach. I want it in the style of, I don't know, Italian Barocca. And I want this high. And it will generate plans for you. It will give you all the materials that you need, price them, and you will understand, okay, it costs this much. Now let me work backwards on that.

• 47:04Craig S. Smith?

  Wow, that's how far do you think we are from that kind of a integrated solution?

• 47:11Speaker 01

  Internally in the company, I'm known as an AI optimist. So I just stay away from this. I would say knowing construction is at least 10 years.

• 47:22Speaker 01

  Even if there is AGI and ASI, since I still think this is a harder problem to solve.

• 47:29Craig S. Smith?

  Yeah. And in pre-construction, is that when you look at alternative materials? Or is all of that worked out in pre-construction, right?

• 47:39Speaker 01

  Yes, you also select who gets to build it. They give you their bit. I will build this with these materials. I will build with these materials and you decide.

• 47:49Craig S. Smith?

  Yeah. And how much, there's so much happening in new materials. Does Kogel have in its training data access to to to new materials or new methods to recommend? Or is that not really what Togol is intended for?

• 48:10Speaker 01

  I think Togel has access to the latest large language models. We rely on Gemini, for example, for our reasoning layers. And Gemini itself has scouted all of the internet for all of the possible nuances and latest discoveries. And so in that sense, yes, we have understanding for what is out there, even if there's some cutting edge research into new materials. If you really want to, you can go into Togel and use the large language models to investigate how you really can improve the efficiency or whatever is it that you need to improve in this building. And I think in that sense, yes, we do.

• 48:49Speaker 01

  We do not advertise it, advertise Togel as the cutting edge optimization utility for value engineering. Value engineering is like the last stage of pre-construction. So we'll get there, but there's many more problems to solve for that.

• 49:09Craig S. Smith?

  Yeah. And in your job, you have a product. Are you moving along a roadmap toward this larger vision? Or are you mostly dealing with customer specific issues?

• 49:25Speaker 01

  It's a great question. It's a constant struggle, as is with any B2B SaaS. We are trying to say no to some customer requests if they are not aligning to the bigger vision that the digital find.

• 49:42Speaker 01

  In a sense, I think we are moving to the big vision, but it's not like we're only building, building and not having any customer validation. We know what are the stepping stones to that vision and we're methodologically trying to build through those steps.

• 49:59Craig S. Smith?

  Yeah. Okay, I'm running low on questions. Is there anything we didn't cover that you wanna talk about?

• 50:08Speaker 01

  No, I think that was great. I'm I'm happy I got to talk about construction, which I'm obviously I'm excited about. I don't think there's anything else to cover.

• 50:21Craig S. Smith?

  Yeah, okay, I'll I'll put this together, I'll send a an edit to you guys that you can look over. If there's anything in there you want cut out or amplified. But yeah, this is interesting. It's AI is attacking these verticals and this is one I haven't really thought about. And I'm just curious on the construction phase, not pre-construction. Is there a lot of AI involved now, kinds of things?

• 50:52Speaker 01

  In construction, for example, as you're building, you want to understand how is the building progressing? So there is startups who have cameras in construction and tell you, this is the state of the construction. These are your plans. This has been done. This has not been done. So it's going to the plan. Be careful because it might go awry. So there's startups who give you observability around that. There's just startups like Procore, the company had done APO recently. They're owning the communication layer, like they are a system of record of construction. As you're starting building, you have to have these different formats communicating to subcontractors.

• 51:36Speaker 01

  That's just like a lot of back and forth documentation. That is less AI, more like software, but that's a super important part of construction.

• 51:49Speaker 01

  There is AI, of course, dexterity, I don't think has been sold yet. But there are robots that help you build and construct things, specific things faster. And then that's like a whole different angle of innovation in construction. There's so many things, this is constantly, Kitvan constantly coming up with new ideas.

• 52:12Craig S. Smith?

  Watching this space, do you think in ten years, as you were saying, will be looking, they'll, they'll be building buildings in a very different way with different materials, largely directed by AI agents?

• 52:29Speaker 01

  Even now, you can do modular buildings, modular design, you can do a lot of stuff, uh, again by robots. I don't think the technology is a limited factor here. I think the limiting factor is again, the people, systems. And I don't think in ten years, much will change. It's the industry that also employs a lot of people.

• 52:52Craig S. Smith?

  Yep.

• 52:53Speaker 01

  That is not going to change in a foreseeable future. Even though there's also these benefits, I can fix a lot of things in my house by asking G P G how to fix them instead of calling the plumber. But the plumber is still needed for a lot of cases and that's the case for the foreseeable future.

• 53:19Craig S. Smith?

  Yeah. Okay, uh, that's, it is Oleg, right?

• 53:24Speaker 01

  Yeah, Oleg Alexander, it's different people call me differently.

• 53:27Craig S. Smith?

  Yeah, but no, this is, I learned a lot, this is really interesting.