“We’re Competing with the Stone Age”: Hungarian Technology Is Mapping the Trees of the World’s Major Cities

By Réka Debreczeni
Originally published by Portfolio.hu on September 4, 2026
Translated from Hungarian

Read the original article in Hungarian

Introduction

This interview explores how greehill grew from a Budapest-based research project into a global urban tree intelligence company working across five continents.

Founders Gábor Goertz and Gyula Fekete discuss the origins of greehill’s technology, the importance of managing trees as critical urban infrastructure, and why cities need consistent, measurable information to protect their existing tree populations in a changing climate.

The conversation also examines how LiDAR, artificial intelligence, and repeated city-scale assessments help urban forestry teams move from reactive tree care toward proactive, long-term management.

A technology company that started out in Budapest now helps cities from Singapore to São Paulo survey their trees — except, as it happens, in the Hungarian capital itself. The reason is a lack of funding, though promising negotiations have now begun: according to the founders, if the expertise available here is paired with an adequate budget, Hungary could soon join the world’s leading ranks in the management of green infrastructure. The shift comes at just the right moment: according to greehill’s CEO, if we don’t start spending on urban trees now, climate change means there will be nothing left to save within five years. An interview with founders Gábor Goertz and Gyula Fekete.

If you had one minute to explain what greehill is and what you do, what would you say?

Gábor Goertz: We help cities make their urban tree stock healthier and safer, and make sure the positive impact of urban trees on the population is as large as possible.

Trees play a very important role in urban life — what we call ecosystem services. They provide shade, lower temperatures, capture carbon dioxide, filter fine particles out of the air, and have many other positive effects on the people living around them. Even nearby property prices go up: the more trees there are, the higher property prices tend to be.

The climate is changing at a brutal pace — just look at what’s happened in the last two years — faster than trees are able to adapt. Today, 50 percent of trees already live in a climate zone they are not prepared for, and by 2030 that share could reach 75 percent. It’s too hot, it’s too dry. The result is that more and more trees fall ill or die, which reduces their positive impact even further. On top of that, there will be more and more tree-related accidents, endangering residents and infrastructure.

The two founders: Gábor Goertz (left) and Gyula Fekete (right). Source: greehill.

Either we wait for all this to happen and suffer the consequences, or we accept that it will happen unless we act. This is what we help cities with: using LiDAR and artificial intelligence, we survey the current condition of the tree stock, and then help decide what the next steps should be. (LiDAR is a laser-based remote sensing device similar to radar — ed.) Which tree doesn’t need attention because it’s in good condition, and which one is starting to lose vitality and therefore already needs attention.

This is a bit like medicine: the earlier a doctor detects a disease, the better they can treat it. But we go one step further. We rescan every two or three years, and we don’t just help treat diseases — we help actively manage health. Just as we don’t only go to the doctor once we’re already sick, but also go for regular check-ups.

Just as medicine is moving from reactive disease treatment toward the active management of health, we are applying the same shift to the management of the urban tree stock and urban green infrastructure.

Where did the idea come from to turn this into a business?

Gyula Fekete: We already surveyed Budapest using high-resolution LiDAR technology back in 2014. Word of that spread, and that’s how we ended up in Singapore, where we were also working on a detailed survey of the city, though not for tree-management purposes.

The Singaporean authority there — roughly the equivalent of Hungary’s FÖRI [Forestry Directorate] or the Lechner Knowledge Center — commissioned us. They introduced us to the local parks authority, NParks. The idea came from them: since such a detailed point cloud was already available, couldn’t the trees be automatically extracted and evaluated from it?

They had three million trees and about three hundred people. They essentially never reached the end of the job — they kept measuring the trees continuously instead of actually being able to manage them.

That’s where the conversation started. We thought the problem could probably be solved within a year, but it needed research first. That turned into a research and development project within a local consortium. It included the Singapore space agency, SZTAKI [the Hungarian Institute for Computer Science and Control], the land registry office, the university, NParks, and us. By that point we already had enough of a track record that they believed we might actually be able to do it. We ran this as a separate small company, because Budapest Közút [Budapest Roads] couldn’t operate there. (Gyula previously worked at Budapest Közút — ed.) Around the time this R&D project started, I met Gábor. Gábor happened to have some time, and I needed a bit of extra help. That’s how the threads came together.

In the end it became a two-year R&D project, which went so well that NParks said: if we could actually deliver this in practice, they wanted us to carry out the work in Singapore on an operational basis under a larger, five-year framework agreement. That’s how our story began.

Incidentally, Budapest Közút has been measuring roads using this same core technology — mobile laser scanning — for road-management purposes since 2013–2014, the first in the world to do so. A survey like that generates a huge amount of information, from which you can build a very precise, standardized record of the condition of the roads.

I was in charge of that program in 2014. Word of it spread around the world, because at city scale this technology was a novelty at the time. That’s how we ended up in Singapore as well. There, Közút itself, as a state body, couldn’t operate, so we set up a separate company.

The original goal of the Singapore project was likewise to survey the city in detail using mobile laser scanning. Green areas weren’t a separate target area at first, but they became a very important element for the Singaporeans, and that’s where the demand surfaced.

The device used to survey the trees. Source: greehill.

Gábor Goertz: So the demand itself came from the client. Our client, NParks, is one of the most modern and progressive green-space management organizations in the world. It’s characteristic of the whole Singaporean mindset that they take risks and innovate. Even back then we had very specialized expertise, which is why we won the R&D program, and NParks defined exactly what data we needed to extract in order for them to manage the city’s trees better.

That was a special situation, because the client provided not just the funding but also the specification. The project went well, and then we won a five-year tender. Under that, we built a digital platform which today is the digital backbone of tree management in Singapore. That first contract ended last year, and now a new, ten-year period is beginning. Singapore was our first client, and to this day it’s one of our biggest.

Gyula Fekete: Most startups start with an idea and then try to sell it. For us it was the opposite. It was a very lucky coincidence — the market came to us.

Gábor Goertz: We also see it a bit differently in terms of whether this is really a niche market, because in fact this is a global problem. Not just in Hungary or Budapest — right now every city is struggling with a changing climate, which means trees need to be managed with better, more modern methods.

Before greehill, this basically worked in one of two ways: a city either knew almost nothing about its tree stock, or people — arborists — walked through the trees over the course of ten years and manually assessed their condition. But that way, the records were never truly up to date, because by the time they reached the end, they had to start over.

In São Paulo, for example, we won a project after a massive storm. 1.2 million people were left without power for four days because so many trees fell onto power lines. São Paulo is a city of 18 million, and the mayor said this couldn’t continue. We got the job within four months. We’re now nearly at the finish line: we’ve surveyed the whole city, and a detailed tree registry has been created, which they can use to manage the tree stock.

In India, three months ago in Mumbai, a tree fell on a school bus and an 11-year-old child died.

Unfortunately, we often see that politics only reacts once a tragedy like that has already happened.

That’s also why it’s a big problem when people say — as Mr. Bardóczi [Budapest’s chief landscape architect — ed.] or Péter Dezsényi [head of Főkert, Budapest’s public parks company — ed.] have both mentioned — that a city the size of Budapest doesn’t have the money for this. I don’t think “there’s no money” is the right way to put it — the real issue is that the money is being spent on other things.

When we were negotiating with New York, the manager said unfortunately they didn’t have the money. I asked what that meant — what New York’s annual budget was. 128 billion dollars. Our project would have been a sub-project, four million dollars. In cases like that, you can’t simply say there’s no money — it’s really a matter of spending it on other things instead. I think the same applies to Budapest or any other city.

If we don’t act now, these cities will become unlivable within five years.

I was in Dallas, where it was 48 degrees Celsius downtown, and there were no trees either. You can’t get around normally. Budapest lives off tourism — it’s one of the most beautiful cities in the world. I can’t imagine simply accepting this process. After the big 2022 heatwave, by 2023 the quality of the tree stock had already deteriorated to the point that it’s visible to the naked eye if you walk through the city today. If we don’t start addressing the problem, this won’t be a “nice to have” — in three or four years the city will become unlivable. That’s bad for us as residents, and bad for tourism too.

At the same time, we’re seeing growing openness worldwide to the idea that trees aren’t decoration, but a significant element of the urban ecosystem. If the tree stock is in poor condition, people’s quality of life and safety both decline.

In this context, big announcements are often about how many new trees a city has planted. How good a metric is that really?

Gábor Goertz: A tree needs 15 to 20 years to reach maturity. That’s why we’re advocates of preserving existing trees. Planting is important, of course, but only about a third of the trees planted today will survive. A newly planted tree needs 15 to 20 years before it provides the same ecosystem services as a mature tree. Until then, it doesn’t give the same shade or the same cooling effect. So we need to invest simultaneously in preserving the existing stock and in planting new trees. But the first and most important task is preserving what we already have.

Gyula Fekete: We’re hopeful that more and more politicians won’t wait for a tragedy, the way it happened in São Paulo or Mumbai. In Hungary, at least there are now public statements about this — before, even those were rare. You could mention Mayor Karácsony or Minister Vitézy — they’re already talking about it. We’re hopeful that the statements will be followed by the first concrete steps.

It’s a bit like someone wanting to repaint their apartment: first they need to measure how big the apartment is and how much paint they’ll need. The same is true for trees. Until I know how many trees I have and what condition they’re in, it’s very hard to make any strategic or operational decision. That’s typically why we represent the first step toward a city later being able to act proactively. This opening has already begun elsewhere, and I’m hopeful it will increasingly happen in Hungary too.

What kind of solutions are you currently competing with?

Gyula Fekete: The Stone Age. That’s exactly why we’re already present in 130 cities. More and more cities are recognizing that they have a choice: spend five to ten years manually surveying their tree stock, or bring in greehill, which can survey the whole city in a few months and deliver an up-to-date registry within half a year.

Gábor Goertz: Today, greehill doesn’t really have a true alternative. That doesn’t mean there won’t be one in five years, though. We didn’t just start working with artificial intelligence recently — we’ve been developing our own AI since 2017. We don’t use generative AI; we use a different approach, our own technology, because we need to build robust systems. Even ChatGPT states outright that artificial intelligence can be wrong.

We, on the other hand, can’t afford to be wrong. If we’re analyzing at what wind speed a tree will fail or fall, and we get that wrong, and the tree then falls on someone, that becomes a very serious problem for us.

That’s why we’ve developed a fully proprietary technology that we know for certain delivers reliable results. In the specific task we perform, generative AI technology currently cannot achieve better results. The approach we use is called discriminative artificial intelligence.

Modeling a tree’s trunk stability under different wind strengths. Source: greehill.

Gyula Fekete: With generative AI, if you show it several pictures of cats, it can generate a new cat that doesn’t exactly resemble any of the originals. Discriminative AI, which is what we work with, does something different. It sees, say, five pictures each containing a cat, and then, if we give it another picture, it can recognize that there’s a cat in it. It doesn’t add anything new — based on the input it receives and the patterns we’ve trained it on, it recognizes what it’s looking at. We train it on trees, so that it can recognize them.

Gábor Goertz: In recognizing tree species, our system today works with about 90 percent accuracy. We fed the same data into a generative AI system, and it achieved 23 to 30 percent accuracy. This represents many years of work — altogether we’ve invested more than 25 million euros in research and development.

Since we’re talking about Singapore — you said they apply the most innovative green-space management methods there. What could be learned from them?

Gyula Fekete: For example, part of everyday operations there is that tree-care workers don’t head out into the field right away. First they look at the trees virtually, in our system. They review in advance, from the office, whether it’s really worth going out to a given tree, and if so, they arrive on-site already prepared.

They also use a huge number of sensors, and tree planting happens on an extraordinary scale. Part of their philosophy is planting a very diverse flora — many species that belong to the original jungle. So there are far more tree species in the city than in a typical European city. The advantage of this is that the ecosystem becomes more resilient. It can respond in many different ways if a disease or some other problem appears.

Gábor Goertz: The management of urban trees has actually developed a lot in Europe too. In Hungary, for example, there’s a great deal of expertise in this area specifically. Singapore has both expertise and money. Hungary has a great deal of expertise specifically. We work all over the world, so we can compare the level of readiness across different countries.

I think that if the existing expertise in Hungary were combined with an adequate budget, we could certainly move into Europe’s leading ranks within a foreseeable time.

Tree management has a long history here. The knowledge and experience are there — what’s mainly needed now is the money to turn all of that into truly world-class operations.

There really isn’t much money, and Budapest doesn’t use this technology either, even though the company started here. And yet your headquarters is here. Why did you decide to stay?

Gábor Goertz: Let’s start with the fact that we both live here.

Gyula Fekete: This is our home country.

Gábor Goertz: Yes. My mother is Hungarian, my father is German. I was born in Germany, grew up in Paris, and moved here 22 years ago. My children were born here, so now this is my home country too. Beyond that, you can still find very good engineers here. Not just Hungarian engineers — Budapest as a city is internationally attractive too. We have French, German, Ukrainian, Russian, and Kenyan colleagues working with us. We’re a kind of melting pot. What matters to us is that someone is professionally strong and has the right mindset.

What we do is interesting from that angle too — I don’t want to downplay anything, but it’s a different thing to write banking software than to take part in a project that can have a positive impact on the lives of a significant part of humanity by helping better manage and protect urban trees.

I find it very inspiring that we’re now working with Mumbai, working with Australia, and that Japan has reached out to us too, because they’re very interested in what we do. We’re engineers, we get to work with the latest technology at an international level, and at the same time we’re doing something that actually makes sense.

Gyula Fekete: Hungary is a good place, and we’re here. It’s not just R&D that happens here — a significant part of our operational work is actually organized from this office too.

I don’t want to move anywhere else, and I don’t think either of us wants to be traveling constantly from Monday morning to Friday evening.

What did it take to scale a solution that already worked in Singapore to this many cities? What difficulties did you run into as you grew?

Gyula Fekete: How much time do we have? Do we have two days?

Gábor Goertz: One of the most important things was that we were very lucky with our investors. We found an international investor group, but not institutional investors — very wealthy private individuals instead. People who obviously also want to make money, but who also care about creating something good. These are people who either believe in a project or they don’t. And because it’s their own money, they don’t necessarily think that if everything isn’t perfect after three years, they should pull out.

We sell to the public sector, where the sales cycle is very long. From the first meeting to actually generating revenue from a client takes an average of a year and a half to two years. That period has to be financed. There are extreme examples — São Paulo came together in four months. San Francisco, on the other hand, we’re about to start scanning in two weeks, even though we started working with them four years ago. We won the tender three years ago, and it then took three more years to get through the entire bureaucratic process.

So business with the public sector is very stable, but slow. Investors generally don’t like that. We, however, were very lucky to find an investor group that understands this, supports it, and thinks big, globally.

Basically, we did everything against the textbook. Investors usually say: start in a small market, move step by step, learn, and then expand. Instead, we went straight out into the whole world, and we’re learning simultaneously on five continents.

That’s expensive, it costs a lot of money, but thanks to it we’re already present on all five continents, and in many places we’ve become the benchmark. We’re present in the United States, South America, Europe, Australia, New Zealand, and now we’re building up India. And of course there’s Singapore.

So the main difficulty is clearly that doing business with the public sector is very slow. At the same time, awareness is starting to form that yes, money needs to be invested in urban trees, because if we don’t do it now, there will be nothing left to save later. Whoever doesn’t understand that today will end up paying five times as much later. More and more politicians are starting to recognize this.

Gyula Fekete: I can think of three more major difficulties — two technical and one human. The first technical challenge is scale. Even though we’d already done something similar in Budapest, surveying a hundred kilometers is a completely different thing from surveying several thousand, or even tens of thousands, of kilometers of urban road network. That brought a huge number of technical problems. Every single day we generate 4 to 5 terabytes of data. That has to be delivered somewhere and processed, again and again.

The other major technical challenge is automation. The machine-learning models need to perform well so that trees can be evaluated automatically with sufficient accuracy. The more we can automate, the more cost-effectively we can operate the system as a service. That’s actually one of the keys to our business model. It remains a huge challenge to this day, and our developers spend a great deal of time on it.

And then there’s the human, or organizational, challenge: covering the entire world with a still not very large team, with a service that involves an enormous amount of logistics, while our clients operate the way Gábor just described.

Gábor Goertz: We’re now fully transforming greehill into an AI-driven company. This year, we’re not just using artificial intelligence in our product — we’re shifting our entire operations in that direction too. Not because we want to cut staff, but because we want to move faster. We’re creating a new product category in the world right now, and wherever there’s a market, sooner or later there will be competition too. Right now we probably have a two- to three-year lead, though in this world that’s getting very hard to say with any certainty.

I recently spoke with an executive at one of the world’s biggest IT companies. I asked what their long-term AI strategy was. He said he couldn’t tell me. I asked why — was it a secret? He said: “No. It’s because there isn’t one. We have a twelve-month strategy, because we can see twelve months ahead. Beyond that, we can’t.” And this is one of the biggest IT companies in the world, whose products everyone uses every single day.

That’s why we’re now transforming our entire operation to be AI-driven. We want to achieve more, faster, with the same headcount. We have a global market ahead of us that could represent several hundred million euros a year. We’re ahead right now, but we want to be present in these markets even faster and more deeply.

What does it actually mean in practice for the whole company to become AI-driven? A lot of people use that as a buzzword these days — where does it specifically show up at your company?

Gábor Goertz: We recently held an internal, three-day workshop for the company, with colleagues coming in from all over the world. There were about eighty of us. There was a presentation where they asked: who uses artificial intelligence? Everyone raised their hand. Who prompts? Almost everyone. Who uses Copilot? Well, maybe ten percent. Who uses AI agents? Two percent.

We’re now transforming the whole company’s operations. This change has to be carried through with every single colleague. The idea is that every colleague becomes a kind of manager: they define what needs to get done, and the AI agent acts like the subordinate that carries out the routine tasks.

This is a real transition. It’s not a matter of saying, “Okay everyone, go ahead and use Claude.” We’re working together with an outside firm, and we’re reshaping our entire operations, our processes, and our way of thinking in this direction.

Gyula Fekete: Managers are no exception either. We’ll be using AI agents in our own work too. For example, to review how we’re doing with our emails, or to help prioritize our tasks. Everyone will build these kinds of small systems for their own area. The goal is for everyone to use AI agents at a minimum. That’s already two levels above simply typing in “where’s the nearest place to get raspberry syrup.”

Gábor Goertz: Our goal is to increase productivity by 30 to 100 percent with our current headcount. In a year, we’ll be able to say whether we succeeded.

We’ve already talked about the public sector — the sales process is long, and politicians often only start treating this as a priority once some kind of accident has already happened. But what are the upsides of having the public sector as your client?

Gyula Fekete: In Hungary right now, what I see as a positive is that interest has started to emerge even at a very high level.

Negotiations are still ongoing for now, but over the past three months intensive conversations have started on several fronts, clearly coming from the political side.

The other advantage is that once you’re in, you’re in.

Gábor Goertz: Exactly. We don’t just carry out a one-off survey — we essentially change the way cities manage their green assets, their green infrastructure.

The beauty of it is that it’s very hard to get in at first, but once you’re in, it can turn into a long-term partnership. In Rome, for example, as part of a larger pilot project, we scanned 84,000 trees, and next year we’ll do the whole city. In Singapore, there was first a research project, then a five-year contract, and now I’ve just signed a ten-year contract. So these are very stable, reliable partners.

Singapore is your oldest project. I imagine the survey has already been repeated there several times.

Gábor Goertz: Some areas are surveyed annually, others every two years.

How much more can you tell once you have more than a single snapshot, but several surveys instead? At what point does this really become a valuable time series?

Gábor Goertz: Even the first survey provides a huge amount of information. But with the second one, we can already see how the trees have developed. It’s a bit like with children. When a child is born, there are growth curves — for example, how tall a boy is typically at six months, then at nine months, and so on. Parents measure this continuously.

Something similar can be done with trees. We can see, for example, which tree is growing more slowly than the group of comparable trees around it.

But there was also a tree that grew much faster than all the others. It turned out there was a burst pipe underground next to it, so it was getting so much water that it was thriving.

So there are both negative and positive outliers. These trends only become visible once you have multiple surveys, while current data is also continuously available. This, again, is like a medical check-up. I can go to the doctor once, or I can go every two years. I, for example, have a value that shows my kidneys aren’t functioning as well as they should. Whenever a new doctor sees it, they’re always alarmed at first, and then it turns out it’s stable. That’s just how I was born.

In cases like that, the doctor looks at the trend too. If it’s stable, there’s nothing to do about it. The same is true with repeated surveys: I always have a current picture, and from the trends I can see whether something needs attention or not.

Gyula Fekete: From the trends, we can also identify which tree species are underperforming or overperforming. That’s important, because for new plantings it makes sense to prioritize species that perform well in the new climate.

Let me give a hypothetical example: suppose chestnut trees are performing very poorly, while lindens are doing well. In that case, it would make more sense to plant lindens going forward, because chestnuts are expected to die out. This kind of information can be seen from repeated surveys.

Gábor Goertz: There’s one more important thing. Every city has goals it wants to achieve. Regular measurement lets them actually prove whether they’re moving in the right direction, and also what happened with taxpayers’ money. Doing a single survey is one thing. But if a city says it’s investing amount X in expanding the area covered by tree canopy, or in improving the average vitality of the trees, then regular measurement lets us show whether it’s really moving in the right direction.

It’s very simple. If I have a project where I want to get from A to B, first I need to determine where A is and where I want to end up — that is, where B is. Then I keep measuring, to see whether I’m heading in the right direction or need to correct course. We do the same thing.

Do you have a favorite project, or a city where you especially enjoyed working?

Gyula Fekete: Singapore is a lot of fun, but also a challenging client, and that’s exactly why I like it. I’ve learned an enormous amount from them. And of course I also love the cities back home.

Gábor Goertz: There are smaller cities with twenty or thirty thousand trees, and there are cities with several million trees. One of the most beautiful things for us is that everywhere we go, we meet people who love their profession. Wherever we go in the world, we meet people who want to do good. It feels good to work in this field. Of course it’s also about money, but not only about that. The people working on this put their heart into it. We often see that even under almost impossible conditions, they get the maximum out of what they have available, because they genuinely care. They care about the trees, and about the people living in their own communities.