Julian: Hey everyone. Thank you so much for joining the Behind Company Lines podcast. Today we have Marco Giardina, co-founder and CTO of Geowox. He Walks provides an automated property valuations for mortgage providers. So fascinated to, to get into your background, especially because it's so data intense and, and really excited to learn more about.

What capabilities founders can, can take away from the data that they're, you know, either keeping track of, not keeping track of, and, and really just hear your experience, Marco, on the data side and as well as, as what you're building and, and how you got involved into it. And and also, you know, the, the different ways companies can utilize your information to better, you know prepare themselves or, or take insights from the information that, that your company gathers.

So before we get into all that, Marco, what were you doing before you started?  

Marco: This is an interesting question. First of all, thank you, you and having me here today on your podcast. I really appreciate this opportunity. So, before jokes, I was I spent, let's say my career as a data scientist in banks.

Telecommunication companies and media companies. The kind of main objective was to acquiring more customers, selling more products of services to them, and try to identify. The customer to work worth to retain. So that somehow is what you do in any company, any industry as a data scientist. So it's what everyone asking you to do.

Julian: Yeah. Yeah. What's that process like when, when a company sets up, say, do all companies set up their data to be taken in the same, or are there ways that they need a data scientist to not only. Create the intakes of the information that they need to collect, but also interpret it. What's the process like?

Are most companies set up or, or do they kind of fall short in some areas and aren't collecting everything that they need?  

Marco: This is a great question. So based on my experience so I started in late 2000. 11 as a data scientist. And at that time was kind of the beginning of this new era of using data to drive, let's say sales.

And usually when every time I went in a new company you need to work with what you have, try to figure out where the were stored, how they were organized, the. I call this kind of whispered knowledge on how things work, who was in charge, and usually due to the turnover was always challenging to understand who did what, why was in that way.

And every time something let's say, Breaks, you are kind of worried what to do. So, but what I will say is most of the time you work with what you have, try to make it better, try to simplify it. But there is a lot of kind of Work to be done. There is a lot of politics involved, but in how things are established, especially in a corporation.

Julian: Yeah. Well you mentioned you know, it was kind of a new age in using data to drive sales. Was sales the primary goal for a lot of these companies or. Were there other ways that then they started to figure out they could use data to maybe have insights in their business or their spending or, or the, the products that they should actually kind of invest more time and, and resources to What were the different ways that data was then evolving to be used for?

Marco: I think the, the first goal was always to understand their own customers. So who actually the customer were, what product they, they have, how much they use this product, what was let's say. The factor or the drives to sell more or what triggered them to, to churn. So for instance very simple example over communicating to customer bring them to opt out for marketing communication, you lose opportunity to sell.

So this seems very intuitive. But when, let's say you do at scale and you work, each team work in a silo, they don't realize what the marketing team does against what somebody else do in the same organization.  

Julian: Yeah. Yeah. And, and is that kind of a, a silo from the way data is communicated across teams or is.

Or are we missing pieces of data that, that isn't being collected?  

Marco: I think I'll say usually is kind of a transformation in terms of mindset because you go from business knowledge, what you know, what you think he's right. And most of the process mirror. This business knowledge to something that a machine tells you to do so, and there is always this.

I did, I dunno, this campaign for the last 10 years, I did this type of activity for the last 20 years so I know what I'm doing. So I think there is an educational process and especially when you work in establish the companies, the good of startups, that you create everything from scratch. You basically create the paradise of let's say any data person because you have everything nice structured.

Yeah. And. Ready to go. At least you start with that objective. Then at the beginning is all messy, everything breaks, and then over time you come to realize what the complexity is good to start the business, but then you need to spend time to simplify it and and make as simple as possible. That is kind of a difficult job to do.

Julian: Yeah. Yeah. How were companies doing that before they had data scientist?  

Marco: I think it was a, I dunno, to be honest, it was a, but at that time, most data were not available. So I think intuition was the main driver and then the other things was, You were not able to, to assess almost in real time if you were doing something good or bad.

So there was a delay in your action, in understanding of your action, and maybe potentially because you were doing so many things at then basically you got always a positive results.  

Julian: Yeah. Yeah. What are the tools that companies use to collect data and then. To store it and, and then to interpret it,  

Marco: In my kind of experience, mainly sql.

So you have a SQL database? Yeah. You have a lot of process done in sql. You would be surprised on how many organization rely on Excel still and nowadays. And and I think there is also, let's say something to keep in consideration, most of the workforce is, has been in this company for a long time, is difficult to upscale, let's say people to use a more technical let's say tool or to change the way of working.

I think it's a generic rational thing, step by step. It became more. Easy to do. Also, the adoption, how even the recruitment is done for this type of let's say professionally is different. When I got my first job was about you know, what is your background? Are you able to. I understand the fundamentals or what you're doing it now, it's more technical.

We have assignments, you have kind of real projects to do. Mm-hmm. To test what you like. So I think we are in a more mature space now, even for I, yeah. Yeah.  

Julian: What, what are some tools that companies use to store the data? Is that SQL as well? Are they, is that like the aws, those types of companies?  

Marco: Aws. AWS I think maybe is the most common resource for that.

So you have different way from storing profiles in I three bucket that is, let's say cheap and functional, and then you build from the, and then depends on what type of data, what is, let's say. Your industry about in terms of security. So there is a lot of things to keep in in mind when you pick, let's say the architecture, the infrastructure that you want to basically to deploy for your product services.

Julian: Yeah. And in terms of implementa, or excuse me analyzation and, and then use of that data, is it one or two main platforms like it, you know, like aws I'm assuming is the most popular, most common. Are there a most common platform that companies use to interpret the information that they gather? Or is it, is it built in within that system?

Marco: I think you have different tools. It depends on preference. I will say also in terms company preference, but also personal preference. For instance you have a sequel that kind of the, the basics. Then you have Tableau, that is a data visualization tool was, let's say one of the first in becoming extremely popular across the organization.

Is easy to use is a drug and drop is a kind of an Excel pilot table. But With be to create beautiful and elegant visualization dynamics. Easy to, let's say, to to move around and and then you have a python. Or at the beginning it was r let's say, kind of programming language for building a machine learning model.

Try to predict future behaviors, some customer clients or I think this was kind of the trilogy. And then with that, so we got the idea of storytelling was kind of a skill not requested in the past. Let's say now a good data scientist should be able to tell a story to make the story visually compelling, to distill the insights in something that is actionable, to engage the audience, to translate in a simple terms for the business, because sometimes you have a very technical things explained in a very technical manner to a business stakeholder that does not understand what we are telling about.

He cannot tell you that he's not understanding it for the role that he plays in the organization. And then you end up in this link book that nobody does anything. Yeah. And the project is kind of shut down. The other person gets piece off because why Everything works. This person doesn't understand. So there is a, a lot, I think Yeah.

Kind of educational process in between. Yeah. nowadays the data person, the engineer is not anymore, only an engineer or another person should be a bit of everything.  

Julian: Yeah. Yeah, it's fascinating. The storytelling component I can see being so valuable. You know, it's, it's like when we, you know, I went to school at a university that was very much research focused and you know, it's one thing to have an hypothesis, it's another to collect, you know, the most accurate data, but then interpreting it and communi.

Where, or the insights that data may have and also may, may not have as well was a huge kind of leap forward in, in the sense that it could guide the next bits of research or the next bits of work that was done on that particular topic. So it's fascinating to see that data is, is in data scientists and engineers are having to take on that responsibility as well.

I'm sure that they kind of can, can, you know, extract some, some teachings or, or some lessons or some, some what is it called? Not evidence, but I always forget the word, but it's like archeologists kind of extract I can't remember the word, but just to extract kind of those, those facts that, that come out of it.

Before we jump, I do wanna jump into Geowox and, and what inspired you to to build a company, but before I, I jump into that real, real quick as With Web three kind of developing, you know, this whole internet computer with, with data and information out being stored and, and kind of a web three structure in the desexualized system is, is, is web three and, and this whole crypto movement, is this changing how data is stored and structured and, and then and then utilized, or is it just another way to do the same thing that you, in the same procedures that you're already doing now?

Marco: I think it can be a. A new way to do the same things, but in a, in a better, more efficient way. Maybe more democratize it the way, because right now not everyone are able to get value out of it. So. It's still the kind of early days to understand out this place in, let's say in the real life because you have a lot of kind of dependency from the, let's say, the old world.

But it is interesting to see how this evolves and how fast you can create value out of, let's say data, because at then these. , everything comes, comes down to how fast you are to interpret what you get. Yeah, yeah. And, and make an action out of it.  

Julian: Yeah. Yeah, that makes sense. Moving into Geowox, what inspired you to, I I, I was looking at your background.

It's not the first company you started and, and was a founder of, but it's the most recent and I'm assuming the most exciting project that. Working on now. But what, what made you go into the insights in terms of you know, property valuations and, and attacking such a, I think a broad field that a lot of not only, I know here in the US we very much value real estate, but I think globally, you know, land and, and assets and asset management as a.

Huge space overall. What made you go in this direction? What inspired you to, to, you know, start building software and the platform and, and things around this type of you know, industry?  

Marco: This is an interesting journey. So everything I think is, is a combination of jokes is the result of a lot of things, but Maybe to explain it simple is being at the right place with the right people at the right time.

So everything started in 2017, basically. So after we, I deployed that location project, the, the mini was. To understand where to do business in which area, and the project was very successful. And the bit of it was, it was easy to explain because it was, okay, I go, in this city, I, there is no competitor.

There is the, I dunno, to open a, a branch very cheap. And there is a lot of good demographics that I can capture. So this, the business understood this. Was very successful. And then from there we say, I started to wondering how. We can leverage this location data and starting to talk with one of my dearest friends.

We say, okay, let's explore together how, what we can build with this. And after doing a bit of discovery, we ended up in looking at the properties, looking at the mortgages, and they looking at the lack of technology. That that they basically involve all the process to get a more gauge from the initial application to the, basically the draw down.

Yeah. And let's say Geowox is the main goal to digitize this evaluation process by making property evaluations simply fast and Right, because right now, The process is is, is not very efficient. There are stakeholders involved, but not able to communicate well what's going on in in the process.

Yeah, and this basically as I impact in terms of turnaround. How fast can you. convert An application and is a very competitive market. So usually, let's say somebody's looking for a mortgage is applying for multiple banks at the same time, the, the fastest gets basically the client in. So it is an interesting space where has not yet been let's say fully embraced by technology, but.

It's now kind of. Growing this appetite to make things better, faster, smoother.  

Julian: Yeah. Yeah. It's so fascinating to, to think about the, the whole process and how a lot of us don't even know where that data comes from or where it's interpreted by these different companies, where, where you're looking to get a mortgage from, or property valuations, at least in the US from what I've learned, are based on.

You know, a, a few different stakeholders. One is, is somebody looking at the property and, and, and seeing structurally how valuable it is, you know, whether there's issues with it or not. And then some people communicate within this database that only, you know, brokerages have access to. And then also to add on the layer that some people use internet data like Zillow or all these other pieces of information that our user base to actually gain those insights and then use those numbers.

So what in ways are, are or what partnerships have you kind of created now that, that you've been working on OX for a while, that you're excited about? And, and how are you giving more access to this inform. And how does that kind of work mechanically within the, within the partners that you've made?

Marco: And this is a great question. So, and as you say, the landscape, the landscape is very fragmented when we talk about property, surprisingly, because it's something that It's very important in the life of everyone maybe is the most important financial decision that each of us will make through their, their life.

And over time we, in Ireland, we were able to build let's say our own. And database of property information. So it's busy where all our r and d goes into it because it's what makes the difference. The majority of the data are, let's say public available, but are difficult to reconcile or to, let's say, Verify because they sit in different place, are not very often up to date with what the property is about.

So there is a lot that goes there in bringing the deal together, reconciling them. Verifying them, announcing them because with what you have in the first place is not enough to build let's say an accurate evaluation process and system. Yeah. Basically it's based on data rather than a human knowledge or that use human knowledge in a, an automated manner.

To come up with strong evidence to justify that value. So I think it's a combination between the greater work that governmental institution are doing in making data more transparent, more open, easily accessible and the work of, let's say, Different data providers, you can think about satellite imagery that now starts to become kind of a commodity or let's say information about the address provided by the address the Companies, or let's say there indeed startups like us put in place to solve these challenging problems.

Julian: Yeah, that's, it's so fascinating. You discussed kind of the way like satellite satellite information and imagery is used now as, as a commodity. I was just chatting with a couple founders and it seems like a lot of different ways information is being opened up to access. Not only private companies, but you know, the interpretation of it and then the accessibility for, for users as well is an exciting process.

And I think it is gonna offer more and more kind of progress in a lot of these different industries that, you know, were, were for better or worse, you archaic in a lot of ways that they were using the information that they. I'm curious, you know, how, how many, how many years did it take to not only build kind of the core product but also get it off the ground and gain traction and, and what were some of the exciting milestones once you did you know, secure your first partnerships?

I don't know if you're able to give us those insights but I would love to learn a little bit more about that kind of process.  

Marco: Yes, of course. So at the beginning, let's say we, we start as three co-founders and we were lucky because. Each of us has its own unique skill set. So it was mainly on the data machine learning bit.

And we have Step is also our CEO was on the engineering part and we, Paul, it was on the sales part. So we were able to basically, going into a client meeting, in a discovery meeting and chat about things and say, okay, we can build this. Yes, we can sell it. How much you pay for it. I say, okay, we can build in this timeframe.

Yes. And we make our first sell sell. So basically we were able to sell from almost day one. At the beginning, we were kind of data solution, let's say reports. The ability to reconcile this information maybe is one of the thing that distinguish us the most. Mm-hmm. And then over time, we built all the process that bring us from a row to clean data that Foster our models and then the workflow that help us to deliver evaluation at scale.

So I think it has been a journey so far, but one, maybe the thing that make us successful was this ability to deliver value from day one. . Also the ability to understand the client's needs and respond to it. Have this willingness to understand their pain and map it out, figure out what to do. And one of the decisions we took from day one was to build something that was good from the.

So kind of playing a long term game. So rather than fixing something, it was not good as a starting point. So now this is paying off, but it was a tough choice to make at the beginning because usually you improve what is the, what is in there, and then you mm-hmm. , you somehow are kind of find difficulties to change it or to rebuild from scratch because you are too much into the process.

So basically every time we challenge ourself how we can make it simple how we can Make it better what we should not do. And and we are also kind of a small team, so less than 10 people. So, but we are able to, let's say, to make a lot of value to satisfy our clients delivering. Let's say high quality results to them all the time.

Julian: Yeah. Yeah. That, that makes sense. And I always like to ask this question to, to kind of gauge where, where things are within the company, but what are some of the biggest challenges that Geowox faces today?  

Marco: I think is mainly market challenges. Now, the, the. It's kind of changing rapidly. We are also seeing what is happening almost everywhere, not all in terms of the economic situation globally.

Yeah. But also the decision that big companies are making today. And we are kind of playing in an industry that is not well known for making speedy decisions. So we are enterprise sales for. Let's say the landing space is a very regulated, so is not a fast easy sell, but we have a kind of the foundation, we have a direct product.

I think this is the right time to let's say innovate in this industry. And property is something that is kind of agnostic in terms of, yeah. Market condition. So because in, let's say in our case we is always important understand to understand the value of a properties. So regardless the condition, because in a, in good time, you want to understand in a kind of a speculative mindset in bedtime.

So you wanted to understand what's the real. Of that thing. Yeah. So is it, we are kind of in, in a, like enough to have these agnostic product.  

Julian: Yeah. Yeah. What's if everything goes well, what's the long term vision for geo?  

Marco: We, so we are working to be the dominant evaluation provider in ar that is our core market.

Of course, after we achieve this milestone, we look with the interest to expand abroad. But first things first is let's say the main goal is the Irish market. Yeah.  

Julian: Yeah. I would like to ask that. I know we're coming. To the end of the episode here, but I always like to ask, you know, for this from my founders or my guests what books or people have influenced you the most at the early in your career or currently now instilled in, in, in your values or, or your strategy or your beliefs?

Marco: These are a great question, so let me think one second. One of the book that I'm reading over and over again is the Almanac of novel. So I really like this book. Every time I read it, I found something different that catch. Me my attention. And it is nicely structured so it works somehow in pillars.

I love it because it's the brave of the, of let's say, each sentence. Is already distilled just to take in to like, to make you think. For instance, there is a, I was kind of preparing for today episode and I got to one phrase that I was using at the beginning of the book. Pick an industry where you can play a long term games with long term people and reading it.

Say This is somehow video, what I'm doing. Happened to me. So it is a very interesting book full of let's say good lesson to read asml, think about.  

Julian: Yeah. Yeah. I love that. I love that. Well, Marco, thank you so much for being on the show and last little bit. I always like to have my guests give us their.

Their are plugs, essentially their website, they're LinkedIn, their Twitters, where can we find you and where it could be a part of the vision of Geo Watson. And if I'm a client, where can I get involved?  

Marco: So you can find us@uhgeo.com and you can find me on LinkedIn, Marco Jordan, and just feel free to reach out.

Julian: Amazing. Well, I hope you enjoyed yourself on the show, Marco. I really appreciate not only your insight early in your career and with data and. Companies kind of transitioned and evolved through using it and the different insights that they're gaining, but also, you know, the way that you're building for a long-term vision.

And, and it's really exciting to see the focus and, and the access that's coming about in this landscape. And then how companies like yourself are you know, taking that access to information, helping other businesses interpret and, and use it for their own success. Again, I hope you enjoyed yourself and thank you so much for being on the show.

Marco: Thank you so much for letting me.  

Julian: Of course.

‍