Buying property in Dubai has always been fast, competitive, and high-stakes — 133 transactions happen every single day. But most buyers are still making one of the biggest financial decisions of their lives based on what an agent tells them. AI is now changing that. Joining me is Christian Kunz, Co-Founder of Prop-AI. Christian, welcome to Wall Street to Mena.
Thank you. Pleasure to be here.
Most people buying properties buy based on their gut, their instinct, their agent. What has AI actually changed about that?
Real estate is the largest asset class in the world. But today it is still trading on opinion — lots of agents calling you giving you information. If you look at financial markets and equities, they have been trading on evidence since the 1980s when Bloomberg introduced the data layer. Why was this not possible in real estate before? Two trends made it happen now. First, land registries getting digitised. Second, AI and compute making it possible to process all this data at scale. We are building an intelligence layer for real estate — like Bloomberg did for equities — so you can start taking decisions on evidence and not on gut.
What does Prop-AI do differently from platforms like Property Finder?
The marketplaces are great for generating leads. What we do is tell you whether a property is worth buying or not — and help you avoid six-figure mistakes. We go one level deeper than just showing you what deals are available. We connect sales transactions, rental contracts, valuations, supply data, and demand data to give you instant insights on valuation and how to underwrite the deal. Then we help you execute and manage the property to actually generate the returns.
Prop-AI holds every transaction, every listing, every valuation. What can you see that nobody else can?
Most people think 60% of the data is out there and easily accessible. That 60% gives you average returns. The alpha is in the remaining 40% — additional data sets, the full context, and making different data sets talk to each other. And not just for one property — for 300,000 properties every day in real time. One example: most platforms tell you gross yield — what rent you get divided by what you paid. What you actually care about is net yield. For that you need to understand service charges and a lot of other things. Or: how many similar properties are currently on the market? Is this price above or below market? How many are competing with you? Nobody was able to answer these questions until now.
Can AI alone solve this — or does it need something underneath it?
AI by itself cannot. If you try it with ChatGPT or any others, the answers are not good. AI is very good at language. It is very bad with numbers — in fact it is very good at lying and making you believe something. It requires a trusted data layer. We built that trusted data layer. We have also introduced an MCP so you can connect it with your cloud to get real data. That is the foundation for anything AI does on top. With a trusted data layer and a system that understands the real estate market, you can generate real value. Without it, AI alone cannot solve the problem.
Are institutional-grade tools now available to everyday buyers?
Yes — this is the big shift. Before, only funds and family offices could afford an analyst team to do all of this analysis. AI is enabling any individual investor to access the same type of information. We have proven that our solutions are trusted by institutional investors like Stake, big family offices, and big developers to underwrite with. And now we are giving it to individual investors to benefit from the same.
AI in real estate globally is projected to grow to nearly $1 trillion by 2029. What is driving that?
One key driver is digitisation of land registries — Dubai has been at the forefront. The second is the increase in compute and the AI layers and trust layers coming together. Because real estate is the largest asset class but also the most inefficient — decision-making is very bad — this combination of having the data, being able to process it, and putting AI judgement layers on top is driving the growth.
What are the numbers actually showing about Dubai's real estate market right now?
Dubai has seen unprecedented growth since Covid. We went from around 40,000 to 50,000 transactions per year before to 220,000 plus last year — more than five times. This year we have seen a correction — we are down around 27% over the full year. But if you compare to 2022, 2023, and previous years, transactions this year despite the decrease are still higher than every other year before. Liquidity in Dubai is still very strong. There is some impact on volume and some impact on prices — but it is still a highly liquid and attractive market.
How does tokenisation benefit from a real-time data and AI layer?
Tokenisation makes it easier to access real estate with smaller amounts. But it does not tell you whether this is a good investment or what you should price it at. Current pricing is done by valuers every three to twelve months. If you want a tokenised real-time market and secondary market, you need this data in real time. With AI and the data layer, you can price the asset in real time, reflect new transactions, and investors can take faster and more informed decisions. That increases liquidity and the number of transactions — because buyers and sellers can find the sweet spot faster.
Thank you so much, Christian, for being here with us.
Thank you. Thank you.
