What's Up with Tech?

How AI Maps the Earth Beneath Our Feet to De-Risk Megaprojects

• Evan Kirstel

Share episode

Interested in being a guest? Email us at admin@evankirstel.com

What if the riskiest part of a megaproject is the part you can’t see? We sit down with Jeremy Suard CEO and co-founder of Exodigo, to unpack how multi-sensor scanning and decision-support AI reveal the utilities and soil conditions that make or break billion-dollar timelines. Instead of relying on incomplete as-builts and sparse borings, Jeremy’s team stitches together an MRI/CT/ultrasound-style approach for the ground beneath our feet—then pairs it with human-in-the-loop engineering so owners get confident maps and actionable choices.

We dig into why underground uncertainty consistently ranks among the top causes of delays and overruns, how field data gets fused into rich 3D models, and why a service model—closer to an AI-enabled engineering firm than a pure software vendor—moves the needle on real projects. Jeremy shares a striking west coast story: a major water main bent six to seven feet off the plan right where a gas line relocation was slated, an invisible detour that would have blown up schedules if not caught early. We also break down the ROI: on a $1B job, each day of delay can cost millions, and contingencies for utilities and geotech often add up to 15–17% before work begins. As Jeremy puts it, owners aren’t buying maps—they’re buying quiet.

You’ll hear how Exodigo is scaling from underground data into engineering services, targeting heavy civil infrastructure like grid expansion, transportation, and data centers. We talk adoption hurdles, the flood of pilots, and why human QA/QC is non-negotiable in risk-intolerant domains. Looking forward, we explore adjacent frontiers—from mining to long-shot space resource exploration—while staying anchored to the near-term goals: better utility detection, smarter foundation design, and fewer conflicts in the field.

If you care about construction technology, infrastructure planning, geotechnical insights, or the practical edge of AI in the built world, this conversation is a must.

Support the show

More at https://linktr.ee/EvanKirstel

SPEAKER_00:

Hey everybody, really fascinating discussion today with the team reinventing how we see underground, making construction projects faster, safer, greener, and saving billions of delays and mistakes. Jeremy, how are you?

SPEAKER_01:

I'm great. How are you doing? Good morning.

SPEAKER_00:

Good to see you. Really intrigued by your mission and vision. Before that, maybe introduce yourself. And how do you describe the big idea exotico at the moment?

SPEAKER_01:

So I'm Jeremy. I'm the CEO and co-founder of Exotico. We're an AI startup really trying to change how infrastructure projects are built. I was born in France, but then I spent about 15 years in Israel and now I'm based out of Miami. And our company is a Silicon Valley startup that leverages technology from the Israeli military. And we're here to change construction.

SPEAKER_00:

Well, it's an amazing mission. And for those who aren't familiar with the world of construction and live and breathe infrastructure, why is figuring out what's underground such a big deal?

SPEAKER_01:

So if you look at all the big projects, uh, those those who have trillions of dollars of funding over decades, they're never on time and they're never on budget. And if you look at the top reasons why, there every project has its own reasons, but there's three that are recurring almost on every project in the entire world. And those three are right-of-way acquisitions, which unfortunately we can't help with. The two others are underground utilities and underground soil conditions. It's called geotech. And the problem is that the teams that are designing the project are designing on lacking information because they don't have good maps of the underground. And then when the construction crews discover the reality, it's too late, and everything leads to a ton of delays and like changing the reality of how the project's going to be built, understanding the design was not right. So all of those unexpected issues have huge repercussions on those projects.

SPEAKER_00:

Amazing. And your scans evidently find 20, 30, 50% more than traditional surveys. Um, how does that actually work without digging?

SPEAKER_01:

So think of it as doing like an MRI, a CT scan, and an ultrasound combined. So we don't invent any new type of sensors at ExotiGo, but uh what really sets us apart is multi-sensing. And we measure about 10 to 20 times more data than any surveying crew in the field right now. We uh use AI models to create the 3D maps, and then we have a team of engineers that are reviewing it and stamping it. Uh so we we're doing what's called an decision support AI, which is the only way I think to really leverage AI in an industry where you need to be perfect. And the results are just way better. So nothing, nothing is 100%, but we're way closer than the industry standard, both on underground utilities and on geotechnical data, which are the soil layers themselves. Amazing.

SPEAKER_00:

And you've already worked with major uh transit agencies and utilities and others. What's been some of the uh lessons learned, uh feedback from those projects?

SPEAKER_01:

That you really need to be part of the project. So it's not enough to deliver a product that's the same for everyone. You really need to fine-tune it for the project. Uh, we understand that more and more. And one of the major lessons that we've learned is also that we're we're now uh broadening our products to be more engineering services. And so what we do now is basically instead of just providing the data, then we also help the client uh understand what to do with it. What are you gonna do? Are you gonna move those utilities out of your way? Are you gonna are you gonna change your bridge foundation because of the soil condition we just showed you? So opening what we call an engineering services branch of Exotico and becoming kind of an AI engineering firm is where we're headed, which means we provide the data, but then we're gonna really help the owners of the project understand what to do with it. And the new wave of AI products that's coming out all the time is really helping us lead the charge there. Uh so now we want to really tap into engineering. And I think that's the major lesson learned is if you want to really change a project, a major billion-dollar project, you need to be part of the project and not just provide data or products that are the same for every client.

SPEAKER_00:

Amazing. And you talk about unlocking a$500 billion market. That's enormous. What where do you go first? What are the uh imminent opportunities?

SPEAKER_01:

So we chose to stay for now with heavy civil infrastructure, which if you think about it, it would be like data centers, power, all the green expansion, and transportation, which is always huge. So we're in all those projects. And the$500 billion market is basically uh civil engineering for heavy civil infrastructure. It's a global numbers. US is typically 20% of that. And uh, we are going to bring more and more AI services that are tapping into design and engineering. Uh, and we start where we're strong. So we provide amazing underground data, and now what we're gonna do is the engineering that's based on that data, uh, like as a first step. Incredible.

SPEAKER_00:

So you mentioned AI, and of course, we're in peak AI excitement hype, but you have a very practical approach to how you use AI. You know, explain how and and why and where you're you're leveraging AI technology and you're solid.

SPEAKER_01:

So you have two choices when you're a tech company. Either you do a product that's exactly the same for everyone, and then it's very scalable, but you can't there's a limit until where you can take it to make an influence on a project. Our approach is a bit more similar to what Palantir is doing in their own world, is doing the full service. So we're acting like an engineering firm, which means we abide by all the standards, we have our own engineering team, which are the top guys in the world. But the difference is that the AI is doing all the drafts. And then you have the humans in the loop that are tweaking it, doing QAQC, ping-ponging a little bit. They have their tools that we build for them. And so you get extremely good quality, uh, like you would get from a team of humans, even better, because you're leveraging AI and a ton more data. But the difference is that we can do more and more of those projects and always have the pro review the maps because they spend little time on every map. So that's like a different type of concept. Uh, like I mentioned, we call it the uh decision support AI. And I believe that's the best way to change industries that just don't have room for mistakes. Like in civil engineering or in infrastructure, every mistake will cost tons of delays, tons, like dozens of millions of dollars wasted, and uh there's just no room for it. So, AI, the problem with VAI is it's never 100%. So either you have a product that's 80% that you just sell, and then you have humans doing the rest, or you do the full service, which we decided to do.

SPEAKER_00:

Amazing. Yeah, speaking of things going wrong, anyone who's been involved in construction, things are always going wrong. You must have some great stories, anecdotes around where your technology caught something that was going wrong or went wrong. Can you share some of those insights?

SPEAKER_01:

Yes, we had a client on the West Coast, a big client, who we're designing based on recent as built. Recent as built are supposed to be the closest to reality. It's basically a contractor, like a year before or two years before, buried new infrastructure and gave the map to the client of, hey, this is what we did, this is what we buried. And uh and we scanned it and we found that it was a major water pipe that was going completely off by like six, seven feet, like doing a weird shape. So instead of going straight, and we pointed out to the client, and what happened in the field is that there was a huge rock right there, and the contractor had to go around it, and they didn't put that in the as belt, so it was right in the area where they wanted to relocate a gas line, and nobody would have thought of that. Like, why would a huge water pipe do such a detour? Um, that's when reality hits the hits the fan, and the contractors have to react, and they don't really um bring that information back.

SPEAKER_00:

Yeah, and you have to, I guess you have to convince project managers in an industry that's very risk-averse uh to trust a new way of underground mapping. Uh how do you how's it going? A lot of pilots.

SPEAKER_01:

Like every owner will want its own pilot, even if we did it with their neighbor. Uh so we see kind of as marketing, it's basically spending dollars to prove the technology, uh, where we tell them give us a project where you already have a map you trust, and we'll scan and show you what we find. We'll find a lot more. So non-stop pilots with more and more owners until the the brand starts to kick in. It's very important.

SPEAKER_00:

What what are what are some of the hardest technology challenges you have to overcome?

SPEAKER_01:

Everything is super hard. The things with the break, like with deep technology, is it's never one bright idea, it's always a ton of small ideas that you add up together. And so when you look at what we do, we build hardware. Uh, we have a lot of software, a lot of AI. We are uh controlling operation in the field, even if now it's not our employees, we're switching to third parties to like partners that are giving us people for the work. Uh, we have to control them. The devices are IoT devices, we have to help them scan, so all of that, and then all the back the process at the back office. It's very complex to do engineering maps, um, and it's very hard. And then we have the innovation part. This year we're really uh starting to go fully commercial with our new geotech products, which are the ground layers, and our new engineering services. So then you have like you're innovating on one hand with we have three different RDs, like RD for innovation, geotech, RD for innovation in AI civil engineering, and then RD that scales all of those, which is the bigger group. Uh everything is very hard, but if it was easy, everybody would do it.

SPEAKER_00:

That's true. Um, speaking of things that aren't easy, infrastructure isn't easy. Budgets are huge, but often very tight and over budget over time. Uh I'm here in Boston and I witnessed firsthand the 20-year Big Dig, probably the biggest underground project in US history, I'm guessing. Uh, how do you frame the ROI to decision makers and financial folks?

SPEAKER_01:

So you can calculate the how much one day of delay costs by take the price of the project, let's say it's a billion dollars, and then 90% of it goes for construction, 10% goes for engineering. So divide 900 million by the number of years they're going to build the projects, you get to a couple millions a day typically on those big projects. So every one day of delay is a couple of millions of dollars, and sometimes a lot more. And so if there was one pipe in the way that needs suddenly needs a relocation and it's unexpected or ground conditions, then you start kicking into those days of delay and it starts adding up. And if you look at how contractor price the uncertainty of utilities, they they put a 7% contingency on the on the billion dollar. And then on Geotech, they sometimes put even 10% contingency. So if you combine those, you have 15 to 17% contingency on underground unexpected issues. That's like 150 million on contingency. And the best case scenario is that it's just a higher price on all the bids. The worst case scenario is that it's not enough, and you end up with a lawsuit for$300 million between the owner and the contractor. That's going to be decided in court who was wrong, who was right. Nobody wins. And uh so the ROI is clear. It's like it's like in they're not buying maps, they're buying quiet. They're buying insurance, basically. They're buying the fact that this will not happen on their project. Um, remove, cross those two reasons out of their three to four reasons that might make the project go wrong. Wow, what a great pitch.

SPEAKER_00:

And so beyond construction and utilities, which you have your hands full with, um, I imagine you could see this technology being a real game changer in other areas, uh, archaeology, the defense, mining, climate research, what other uh future ideas do you have in mind?

SPEAKER_01:

I'm thinking of mining because I think that the world needs uh more and more resources. Uh, and I think it's a very interesting uh area. I also think that if the space era will be in our lifetime, which nobody knows yet, then for sure the first thing that humanity will do is mineral exploration in space. Um, so I do believe that as a long-term roadmap, it's uh we would be very interested in in opening a mining branch.

SPEAKER_00:

Oh, that'd be fantastic. Um gotten huge recognition, a lot of funding. People like Time Magazine, Fast Company are profiling you. So really honored to have you here. What's next? Uh what are you excited about over the next few months and a year or two?

SPEAKER_01:

I'm really excited about uh adding the engineering uh services. We're hiring uh some top engineers now from big projects, opening a small group and like proposing those services to our clients. And uh the combo of now doing utilities, geotech, and the very or like the designs around that data for our clients as a combo, I believe would be extremely powerful. That's what I'm excited about for the next year and really like growing uh more and more in the US. We're exploding where uh our growth is extreme. Um, and I think it's because uh it's a huge industry with huge challenges, and I believe our clients are starting to see the value of what we're offering.

SPEAKER_00:

Fantastic. Well, congratulations on all the success. And uh years to more success in the future, onwards and upwards, or onwards and downwards, uh, probably is more appropriate. Yeah, exactly. All right. Thank you so much, Jeremy. Thanks everyone for listening, watching, sharing the episode. And be sure to check out our TV show, techimpact.tv on business Fox Business and Bloomberg. Thanks, everyone. Thanks, Jeremy. Thank you very much.