April 18, 2023
Yogev Shifman is one of the co-founders of Exodigo, a successful AI company that aims to solve the underground challenge, providing underground risk management services. Before joining Exodigo, he worked at Buildots, another thriving AI company that serves the construction industry.
Prior to his tenure at Buildots, Yogev served as a distinguished officer in the Israeli intelligence corps. He directed multi-national and multi-disciplinary programs for real-time embedded hardware and software systems at one of Israel's top deep tech units. He was recognized with multiple awards for his contributions to the field.
Yogev holds a B.Sc in computer software engineering from Technion - Israel Institute of Technology (the Israeli MIT), and he began his career at Intel Corporation before pursuing an MBA in entrepreneurship and technology at Tel Aviv University.
Julian: Hey everyone. Thankyou so much for joining the Behind Company Lines podcast. Today we have YogevShifman, co-founder of Exodigo, a successful AI company that aims to solve theunderground challenge providing underground risk management services. Yogev, I'mso excited to chat with you and learn more deep about your experience as afounder and kind of where you got.
Go and really this whole kind of AIrevolution, how you're utilizing the technology to, take on some differentchallenges that we might not all be aware of. And so I'm excited to get intothe weeds of what that means in terms of what your problem set that goestackling and how you're able to improve that and help companies to do the same.
So, before we get into all that, whatwere you doing before you started the company?
Yogev: Actually a bit beforeI started the company I was in the Israeli. Myself and the other co-foundershad a long career in the Israeli army. All of us are majors in the army andwork for the intelligence scope on different kind of projects and technologies.
A lot of innovation. And I think that'swhat sparked like the, they want to, to create something new and to solve areal world problem.
Julian: Yeah. Describe some ofthe innovation that comes out of, the army and, and really the initiativeswithout getting to, into too much detail if you're, able to share.
But, you know, I think we, weunderestimate the innovation that does come from, more government kind ofresources and. A really successful team of, of professionals around it. And,and so yeah, describe a little bit about the innovation that was exciting thatyou saw um, during your time there.
Yogev: I think what allowsinnovation, especially in the army, is the, the freedom to fail.
it's not a monetized world. You can faila lot. You only need to succeed once, and that's allow you to try again andagain to solve very hard. And that's a, and also they give responsibility tovery young people. So the, the combination of those two along with, national securitygives you a lot of room for innovation that I think that's the hard part inlike commercial companies, that they can't take the risk cuz everything isresources and money and the Army allows you to develop the skill of taking.
With I gain and I rewards, but also veryrisky areas. And we tried to take that approach when solving a new problem instarting a startup because it's basically the same. Only you need to play withmoney, which real world resources that becomes a bit more difficult to play with,but if you already have inside of you, The, the know-how of, how to take thoseapproaches, what risks to take, what risk not to take.
You can develop real innovation.
Julian: Yeah, and, and whatexperience kind of led you, you're going to the army, you've met yourco-founders there. All you are work working for cloud building software,innovating. And then what made you shift into, working within kind of theconstruction and development and that technology space?
Was there something in particular thatcatalyzed, the want and need and, and the focus to go into something in, inthat, in that
Yogev: domain? I think thebasic of it was we ideate for a few months, a few good months. We actually satin a garage, a startup. With an actual car next to us is we try to solve andfind a real world problem.
We wanted to solve some brick and mortarproblem, something real, something that you can touch, something that you canexplain to everybody very simply, what is the problem? What is the solution?The solution itself is complicated. The market is complicated, but the problemis very easy to underst. Eventually it came up to two ideas.
One in cons in not specificallyconstruction, but any underground risk you can take underground. Mm-hmm. And wewent through the same aspects in medicine, like, uh, looking under the humanbody. So it was the ground or the human body. And we chose the ground justbecause we wanted to fill the market faster in in Sure.
In the medical world, it takes a lot oftime
Julian: yeah. Yeah. And whatin particular was, it sounds like you were, you were kind of uncovering twodifferent problems but using similar type of approaches in terms of thetechnology or, or the, the focus of that approach. And so describe a little bitmore about, what Exodigo Go does in terms of the solution it's giving to a lotof these industries.
Cause I was looking through it and Ithought it was extremely cool, but I'd love to hear from your perspective onwhat you're doing and what the technology allows. I, I guess companies toessentially unearth for lack of a better word.
Yogev: Yeah. No, it's a, it'sa perfect world cuz basically what we do is like taking like mri, CD andultrasound and just aiming for the ground and then use AI to interpret everythingcuz it's too much data.
Yeah. What we do is we take verysensitive sensor. Combine them together with I accuracy. Gps, then cover theentire ground. And it allows us to create a very accurate, high resolutionimage of what's happening on the ground for different use cases. Depends onwhat you wanna find on the ground because not everyone wants the same, but itallows you to get an image of what you need underground, and the AI actuallyhelps to pop.
Yeah, what is
Julian: interesting. Yeah. Andwhat were people relying on beforehand? Without the sensors, without thesophisticated AI to really, not only label the data, I'm sure, but alsointerpret it and kind of, build it into certain models so you can actuallystart accurately seeing, you know, what is under there.
What were they using before?
Yogev: The, the most commonway will be just pen and paper. Maps that says, okay, there's a power line heresomewhere. But there are some more evolved methods to do it, but no concretetechnology to actually try to give them an answer. You rely a lot of what iscalled in the industry as as built or as.
That's basically a documentation ofwhatever change they did in a certain area. Mm-hmm. But it's not a livingdocument. Like it has a lot of problems. It's inaccurate and it's basic becausedocumenting is like the most, like Boeing task, let's say, that nobody wanna doit the end of the project. Yeah. So some companies try to solve that.
But we call it, you can call it asmaybes. And then you can't really rely on it. You need the as is, you need tounderstand what is really there, not just as a document. The document is, isgood. It gives you a lot of metadata about what's going on, which, whichelectricity companies in charge of the area.
But yeah, it doesn't let you know whatyou don't know that is there. Mm-hmm. And that's what we're trying to cover andgive you tools to and services that allows you to get more. Of what actually isunderground.
Julian: And what was theincumbent before? Was it just they didn't have the sensors, they didn't havethe data, or they were relying so much on these, or, or I guess they were, theywere working with the reliability of, of that these pieces of documentationwould be accurate, but ultimately finding that they're not.
What was the incumbent technologybeforehand? I know you said pen and paper, but you
Yogev: know, any sensor. Ithink, I think, I think the. They have some sensors that allows you, you cansee people walk around with come rigid sticks, which are electric sensor thatthey can walk around the street and paint where is the power line.
And they have what they call a gp, whichis a ground penetrating radar that allows you to see some object on the ground.Yeah. But those kind of tools and sensors didn't evolve like other. Mm. What wedid, like allowing use to use cloud computing to analyze those kind of methodsallows you to take an expert with GP with 10 years experience and and replacehim with like, maybe taking even the same person but allowing him to cover likeacres and acres of data.
Yeah. Yeah. Because it's allow you, theautomation allows you greater scale of data and better. Yeah, our sensors cancover every inch of the ground basically. Mm-hmm. So you get a high resolution image,like looking at a 4K image instead of a HD image. And that's allows the, thebig uh, turnover instead of, I call it, when they come with the existing tools,they come to an area and they can cover, and they have some method and they canconnect to cables and they can open the manholes.
It's like a crime scene. Yeah. They'retrying to understand what's going on. Where is the, where is the pipe? I cansee a manhole. What is it? Right. What we do is like mourning you along. It'snot like solving a crime scene. You just cover the entire area and you get theresults. Yeah. That's a disruption in the way they get the information and theythe in the way they can.
On the quality of that information.Yeah.
Julian: And, and describe someof the outcomes. Obviously, I would, I would love to learn some of the, maybeif you, I don't know if you have any stories of. Instances where, you know, ifyou don't have accuracy on that data and you start to say, begin to build, ifyou're not able to really understand what's underneath you, what are somenegative possibilities that you've maybe encountered or, or have heard ofanecdotally?
And then what does that insightobviously kind of mean in terms of the the speed at which you could start allthe progress of the project and start to actually build. So I know twoquestions wrapped into one there, but we'd love to hear your feedback on uh, onboth.
Yogev: Yeah. So speed is theright question.
Cause Yeah. In speed, I'm taking it asperformance. When you think about underground risk, you immediately thinks,okay, someone will blow up a pipe, a gas pipe, or a power pipe. Yeah. Powerline, sorry. And that's a very risky thing that can happen. Okay. But even ifyou didn't hit the power line or didn't hit the gas, It's enough that youdidn't know it was there.
Mm-hmm. So a project will just stop.Think of hundreds and hundreds of people building a metal line somewhere. Theywill just stop because there's a line. Nobody know who it is, it belongs to.They will need to go to the municipality and ask them who, whose line is it?Sure. And they will stop even for a few days.
Those few days can. Hundreds ofthousand, if not millions of dollars, depends on the project. And that cancause a snowball effect that can delay the entire project. Yeah. And, andbecause a construction, sorry, just a construction project is a very methodicproject and it's need and relies on the different steps to happen and happen.
Julian: Yeah. And what was youknow, if there's some kind of centralized database, where is that in? And ifI'm a power company, learning, gonna put down a line and need to make record ofit, or I'm a, commercial development company I'm building and I've added someextra layers underneath in terms of the nearing parts to actually connect withpower systems and things like that.
How do you go to find that informationInitial. Versus what are you able to, in terms of not having to go through thatand just seeing what is, how does that change the paradigm of
Yogev: billing? That's also agood question because there isn't one, there isn't a central database. It'smore than that.
Nobody wanna take the liability. So theydon't like to share information too much with one another? Yeah. Some of themneeds to because of regulation to give some information, but they don'tcollaborate. They try to, and I think there's a big improvement in the way theycollaborate, but still, when some project needs to collect all the differentdocuments Yeah.
And and build a an updated picture ofall the documents they have, it's not even the, as I say, the real data. It'sjust the documentation they have. It can take even a few. And that's becausenobody answered the phone or, or whatever. They just, there, there's a lot ofreasons, but, or they don't have the data or it's pen and paper and they needto digitalize it.
They, they don't always know thereasons, but during those month, you can already get the As is information andstart design based on it. Yeah. So even if the regulation needs you to get allthe information in, Just to let everybody know and take the different riskmitigation methods. You can start doing something, you can save a lot of timeand potentially even lose the entire process.
But that's depends on the if the processis able to evolve. In such manner.
Julian: Yeah. It's amazing tothink about how much the, having the ability to have insights can affect somany different aspects of the project. Not just money, not just speed, but alsohow you build and, and the efficiency of whatever you're building to actually,whether it's, correspond with, with the structure or even just oth other,resource systems, how much better that can overall construction can be.
I, I'm excited to, to learn a little bitmore, and if you don't mind sharing with, Tell us a little bit about thetraction you've had thus far. How many people are using the technology? What'sbeen exciting about, the traction that you've seen so far, and what'sparticularly exciting about the next stage of, of Exodigo and, and where you'reaiming to go?
Yogev: I think the mostexciting thing is, as I said before, because we chose the real problem. Get theinterest of everybody you can see clients wants to hear what we have to say.They want to see people trying to solve the, this problem and that's causing alot of good traction. I can go into detailed numbers, but it allows us to getfood in the door in most places and allow us to keep the right velocity.
As a company Yeah. To, to move fastenough between the different clients and get them fast enough and get themcurious about what we do fast enough. Yeah. Because it's a real pain they haveand they give us a chance to solve it. And then the, we need to bring in theresults and we do. Yeah. And that's what delivering the, the, the results thatthe main target we do, delivering the most accurate results, we.
And in time. And I, I think our main ourmain goal is to improve more and more the scale of the services we provide andto more and more clients. That's our main goal. At the moment.
Julian: Yeah. It's so awesometo think about, the ability that your technology has to just even test it outbecause the problem set is so strong with a lot of these companies and, and,and making those, inaccurate, say estimations or decisions can mean so much to,to just, the length and timeline all the outcomes you mentioned earlier.
I'm curious in terms. The AI portion andthe, the complexity of the data. How much is, is there to be interpreted andhow much better is the, kind of the models in terms of how you're trainingthem, get from project to project? Does it accumulate? Does the, does product,in terms of it's accumulated knowledge become better over time and as many moreprojects and kind of learning in that Yeah.
Yogev: obviously yes, becauseit's an AI based, it's always learned, and that's one of the, the advantages webring without technology Sure. Is that you can learn from each site. Yeah. Andyou gain more knowledge and more data and not relearning every time again andagain. Or depending on if one person was sick or something.
Mm-hmm. Or, or his expertise. It allowsus to evolve the models every time more and more and allow us to find mostsubtleties between the different anomalies. Yeah. Like if now you, you're ableto find between different type of utilities. We are now able to find differenttypes of soils, different types of insider utility going to different types ofdo that utility because they have different manners for every utility.
And the different shapes of theutilities. So every part of any project teaches a lot about what we can donext. And it's also a test bed for anything new we are wanna do. We have theold data and we can just reuse it. It's proprietary data that we use and.Basically helps us move along from project to project and, and evolve fromproject to project.
Julian: Yeah. And, and haveyou ever, or have you had any interest in, say, local governments or, kind of,cities that want to know more about what's being built underneath for kind ofmore abstract planning and in terms of, where they want to invest in terms ofresources? Is that something that you're dealing with now or is that kind of onthe horizon in terms of expectations?
In terms of possibilities?
Yogev: That's of course alsointeresting prospects, but more down the line because you need to see theimmediate value with the engineering companies and the local government anddots and, and basically the owners of those projects are the people with the.Motivation Sure.
To solve this issue and they get themost of the value out of it. It's actually interesting that the earliest youbegin to use us, the more value you'll get in the project. Yeah. Because you'retaking less risks. Yeah. We had one project when they brought us in in thevisibility stage, when they actually choose where to go with the project andthen we scanned the line they wanted to go through and it was busy with a lotof utilities and we scanned the other area.
It was clear, so it allows them to movethe entire project to the other side. But before the project started, if theywould've done it as phase later, it would've cost like billions of dollars tomove. Yeah,
Julian: yeah. And thinkingwhether external or internal. Shifting, shifting gears here, what's some of thebiggest risks that you think in Saudi faces?
Yogev: A risk of what naturewhether
Julian: as a company? Yeah. Asa company in terms of whether it's external, macroeconomic climate or in termsof internal kind of mechanism, mechanisms of scaling. What do you feel is, kindof the, the riskier parts or the first debt that
Yogev: keep you up at? Ithink when we started the company it was it was in better time of the hightech, economic economical system.
Our clients are not that affected.They're a bit affected cuz there's a lot of macroeconomics changes, but theyare less affected. But those micro changes because they, they, the project canbe a 20 year project, a 10 year project. Yeah. They'll last sensitive to those.But a lot of them are government and it's very hard processes.
Our biggest. Challenge is the statusquo. Yeah. Is understanding, I think it's a challenge of every, company. It'snot the competition. There's enough for everyone. Sure. The, the problem is thestatus quo. Yeah. Is getting people and clients that are used to what they do.They used to manage the risk that exists, even though they know they can'tmanage it, they know how to not manage it.
They know how. Nobody will get fired fordoing the right, the, the wrong thing. But if you look at the bigger picture,and that's what they're started to do, because they need to be more effective,they need to cover more projects. They need to do the project with less budgetsand less resources. They understand they need to approve, they need somechange, they need some disruption in that area.
That's what allows us to change thestatus quo. But it's a very hard task. Yeah. To move the status.
Julian: Yeah, if everythinggoes though, what's the long term vision for Exodigo?
Yogev: The long term is thateverybody could use that service as they, as they like and for us to providethem with analysis, like basically be the, the one source for true analysis ofwhat's happening underground.
Yeah, the delivery, it's not, I can'telaborate and it's not settled yet, but understanding how it can reach every,every project that needs it around the world.
Julian: Yeah. Yeah. I lovethat. And I love this next section what I call my founder faq. So I'm going toask you some rapid fire questions and then we'll see where we go.
First thing is I always like to open itup. Um, What's particularly hard about your job?
Yogev: Particularly how my,my job is understanding the values that really resonates with the clients.Yeah. But knowing the limitation of the technology. Yeah. And also the, thefuture limitation, not only the existing ones, but imagining what's gonna limitus and what the glass ceilings we can break.
And what of those we can't. Right. Andwe need to choose the right path and understanding what will really resonatewith the clients. And that's, I think, the hard hardest part for me. Yeah.Because we chose difficult technology, but also a different difficult clientsand market to, to walk in.
So you need to have them both. And it'sactually a bridge between the two. Sure. And I think that's the biggest.
Julian: Yeah. And, and youtalk about future limitations. I'm curious, being that the technology, it onlyseems I can get better and better and better. What do you see as the futurelimitations? Is it more technological?
Is it more based on regulations? Is iton businesses and their outcomes? What have you seen to be or predicted thatwould be your, your limitations as you, as you grow?
Yogev: Even though thetechno, the technology is very difficult, we have a lot of very smart peopleand I believe we will solve the technological problems.
I think the real issue will benavigating throughout the different markets and use cases and regulationsbecause they're used to and build upon the status quo. Yeah. You haveregulation that compares you to get certain things that in a certain way, Just tominimize the risk and we offer a new way.
We can do it the same way. You used todo it and cover it, but it's basically things you can get rid of eventually notthe get rid of, of the existing method, but. Getting rid of the, in theinefficiencies that exist in those areas.
Julian: Yeah, yeah. Thinkingabout as a founder yourself, what's something that you're particularly good atnow that you wish you were better at early on in your career?
Yogev: I think dealing withthe us Sure. And internal politics and the, the rhythm, the things that weregoing on here. I think you need to know when you get into a deal and you getinto a new client, that it will take time even if you have the perfect solutionand you need to navigate it the right way and understand where you can movefaster and when you need to wait.
That things you need to master better.Yeah. And you do it as, as the company moves on and the earliest you can, it'sbetter. Yeah.
Julian: Yeah. Well said. 1, 1,1 thing, just in my curiosity about the, the process that you, you, you kind ofwork within in terms of, other founders that can learn from it being that yourtechnology is disrupting a particular space and, it's probably more, it, it'slike you have a testing phase and then you have this long sale cycle and, andyou really can see the completion of.
Probably, you know, with enough timekind of out, how do you as a founder think about or tackle that partnershipinitially, and how do you make sure that you have a successful outcome beingthat you have. To, focus so much attention onto, how the process is going.Where, a lot of founders have more transactional sales, faster sales cycles, alot more velocity, but it seems like your relationships are a little bit moreintimate.
How have you not only maintained those,but you know, made sure that you do well by your customers and have success ina certain project? What are some things that you've learned from and what aresome things that help you be successful?
Yogev: I think that's a matterof balancing the. Of operation you can do and basically making sure you have ahealthy pipeline throughout the year.
It's, it's true that, as I said, therelationship can take time, but you can have a lot of successes and milestonesthroughout. Yeah. The relationship. So making sure to have successes and notjust wait till the end. Sure. It can be smaller. Project poc. Different evengetting a. Meeting with the client, but sure.
It's neat to have a, an ev some kind ofevolution throughout the process and not just waiting for that silver bullet atthe end. Yeah.
Julian: Yeah. It, it's,that's, that's well said. I think it's underestimated the ability to work onsmaller projects and smaller outcomes, even for a lot of different companieswhen they're building these relationships to even act as proof of concepts oreven, do a proof of concept yourself and, and what that means, those little.
In that sales cycle. That's, that'spretty, that's really, that's really strong advice there. That last little bitis I always like to ask because I love how founders extract knowledge out ofanything that they ingest. What are maybe some books or people that haveinfluenced you the most, whether it was early in your career or not?
Yogev: Specific people that influencedhim. Yeah. I, I don't think I have one name that come to mind. I, I. I do lookat some of the people. Most of them are partners or clients and I can't namethem. But that taught me a lot about the market and how it works. Yeah. And gotme into the business side of things, and I think it's, Taught me a lot abouthow the real world, let's say it Sure.
And not just the, the startups, hightech world works and how to navigate there. And there's a lot of very smartpeople you can interact with and learn a lot from and esp and I think that'sthe biggest maybe the biggest move I have to get good relat. As a, as a foreignin the, in the States Yeah. Is that I really connect with them and just, sitwith them, have a coffee or a nice video chat.
Yeah. But if you really trying tounderstand their problem and their industry and not just selling them what theyneed. Yeah. I know it's been said a lot, but I value the relat. And I neverknow what each connection will bring me. Yeah. And some connection brought me alot without even any initial intent of like selling to that person or sellingto that company, but from building a relationship with him, I was able to moveon.
Julian: Yeah. So basicallyhere, it, it's amazing to hear the, the way that your people, kind of, thementors and other, individuals around even some customers will teach you somuch and, and help connect you to kind of the next. It's almost like this ballof yarn at some point. There's so much that, that encompasses this experiencethat, I think a lot of people maybe not value as much, but when they do can seethe success from a, from a healthy network of individuals around them.
Last little bit is I'd like to make surethat we didn't leave anything on the table. So, yo, is there any question thatI didn't ask you that I should have or anything that you would've liked to chatabout? Anything that we left that you wanted to mention to the audience beforewe leave?
Yogev: Not about the company,but I do wanna emphasize anyone would wanna start a company of solvingsomething need to for himself, making sure it's something real.
He doesn't have to be something big, hejust ha need to solve a real problem to a real person can be in any market andit's need for him to, to be able to explain to himself the differentiator wehave. Sure. From anyone. What it brings to the table as a company and to theworld. What, what added value it brings within.
Julian: Amazing. Yeah. Yeah.It's Yogev. This has been such a pleasure chatting with you. Not only learningabout your early experience and kind of what that taught you and, and how you,started go, but how you focus on customer problems, the relationship you havewith your clients and, and kind of that whole process there.
But also, the technology and how muchinsight you can gather. Based on the sensors that you're building and how much,positive outcomes that that can have. And, and we as consumer or not asconsumers of your product, but as individuals and in society can maybe even seethose outcomes start to, come out in, in our day lives, which is fasterprojects being, completed and, less kind of impact on our lives, but also,maybe even better sustainability outcomes.
So it's amazing to see what yourtechnology can do. And last little. Where can we support you? Give us yourLinkedIn, your websites, where can we, support you as a founder and also go andwhat you're working on?
Yogev: Yeah, you can visitour website, Exodigo.com. You can see everything there. You can even check outjob page.
And you can also follow me on LinkedIn.I try to be active there mainly on also helping young entrepreneurs. Productmanagement insight that can help you get better values and better traction inyour business, like connecting your technology with your business side.
Julian: Amazing. Yogev, it'sbeen such a pleasure chatting with you.
I hope you enjoyed yourself on the show,and thank you so much for being on Behind, Company Lines
Yogev: today. Thank you forhaving me.