“We call it the Tinder of real estate”: This app uses AI to help you house-hunt

AI has done some heavy lifting over the past year—ChatGPT is writing high schoolers’ essays, deepfakes are simulating Joe Biden and Donald Trump playing Minecraft together, and the tech is working toward saving lives. Now, our incoming robot overlords may be taking on their greatest challenge yet: the Toronto real estate market. A new real estate brokerage–cum–tech company, Wahi, is using AI to make three major changes to the house hunting process: it compiles a huge range of information that buyers can access directly, crunches agent info to recommend the best possible person for a given neighbourhood and lets buyers play the part of agent themselves—even earning them a bit of the commission. Here, Wahi founder Benjy Katchen explains what he calls “real estate Tinder,” shares how people are earning money while making massive purchases and breaks down whether the AI onslaught should make homebuyers nervous.

Where did the name Wahi come from? It means “place” in Hawaiian. We thought it was perfect since we’re all about helping customers find their place.

Do you have a place in the city? I’m living in a 120-year-old detached house in the Annex, but I’ve bought and sold a few homes in Toronto. I wished I had more control over the process. For instance, why do I have to call my realtor or their assistant just to set up alerts for when new homes become available, or to ask what a property was sold for? There were all these roadblocks to getting basic information by myself.

Related: Three AI execs explain the tech industry’s extraordinary powers—and its potential dangers

Lots of people find house-hunting infuriating, but not many turn that rage into an app. How did that happen? My background is in digital banking and financial technology. I worked in that industry for 20 years. In September of 2021, I was introduced to some private equity folks who wanted to create a digital platform for all things real estate. So we launched the app in August of the following year.

For those of us who struggle to understand tech and real estate, could you break down how Wahi works? Basically, it’s designed to help you find your next home. It lets you see a property’s past sale prices, school scores and other specifics in an easy-to-read layout. We also have what we call “the Tinder of real estate,” where we match consumers to the best real estate agent for them based on what they’re looking for. Alternatively, we have a discount cash-back program where buyers work with one of our remote realtors and organize their own house viewings with our tour assistants. Then, because the costumer is doing some of the work, they earn 1.5 per cent of the sale price as commission. So, if you bought a $1 million house, you’d get $15,000 back.

Tinder is pretty straightforward, but I’m assuming you’re not talking about swiping left and right on houses. How tech savvy does a potential user need to be? If you know how to use Uber or Airbnb, this should be easy. Of course, my dad is 77 and can’t use Uber, so there may be some people who find it tricky. We could still find them great realtors, though. The layout on the app is a simple map with filters across the top. But we also have the “home hub,” a personalized list of places that fit the requirements users have set up.

What’s really been catching people’s attention is the fact that your app uses AI. Where does that come in? We use AI to deliver the most accurate instant value estimates in Canada by compiling and analyzing all the different attributes of the property. For example, with a condo, we’d incorporate square footage, number of bedrooms and bathrooms, whether it has a balcony, plus recent sales in that building and similar buildings nearby—and all in in real time. Estimates are so complex; you need machine learning to get that quick assessment. It doesn’t take the place of a formal appraisal, but we’ve tested it against other instant estimates and so far it’s been the most accurate.

People are rightfully wary of information provided by algorithms. Should they be concerned when AI is put in charge of real estate? I’m baffled trying to think of how we could have a hidden agenda here. If the app shows people houses they don’t want, then they don’t buy them. This is the most important purchase of their lives. They’re not going to buy a home just because an app puts it in front of them.

But could a glitch could screw up a purchase? There’s potentially millions of dollars on the line. Not really. We’re a licensed brokerage and a member of all the real estate boards across Ontario. There are still humans involved. And of course, there’s still the due diligence that every homeowner should take—going out and looking at homes, getting inspection reports, getting pre-approved for financing. There are no more risks with us than with anyone else.

Related: Meet the U of T student using AI to look for aliens

Are you worried about bad actors trying to hack the process or commit fraud? We may actually be better at protecting consumers than traditional methods. Normally, if you’re trying to find the best realtor in a given area, you’d come across 25 people and have little information to differentiate them from one another. But, when we recommend someone, it’s because they’ve done at least 10 sales in the neighbourhood over the past three years, write good listings and have a higher transaction volume than normal. We have all their stats.

Your website says Wahi gives users the same data and insights that agents have. But what do you provide that people can’t get on Zolo? We’ve mapped out every single school district for every school board in Ontario. Normally, you’d have to type each individual address into the TDSB’s website to see if it’s within the boundaries of the school you want. But, if you click on a school in the app, all the listings in its range just pop up. A week ago, we launched a feature where you draw a shape on the map with your finger and it shows you all the listings in that area. It’s not just the data but the experience we layer on top of it.

Critics have said that, while Wahi’s you-do-the-work option can net you some cash, it’s no substitute for a real-life realtor who can point out cracks in the basement. I think this is where clients want to have their cake and eat it too. If they take the Tinder option, they’ll get a realtor who knows everything, like whether the house has a termite colony or flooding. If they use a desk agent and our tour assistants, then they save some money and get the commission—but they lose that level of local knowledge since they aren’t paying for it. We’re giving people the choice.

Do you ever worry about taking jobs away from traditional realtors? That’s a sensitive topic, but I don’t think most realtors are that productive. Seventy-nine per cent of realtors in the GTA sold fewer than five homes in 2021, and 32 per cent sold none at all. I’ve heard from professional realtors that the masses of part-timers give their profession a bad name by leaving consumers dissatisfied. So there may end up being fewer realtors overall, but I think it’s a net positive.

Do you use AI for anything aside from real estate? I’m planning a trip to Greece, and I like to ask ChatGPT for its hotel and restaurant recommendations and compare them with ones from locals and people I know. ChatGPT doesn’t quite substitute for someone who’s been on the ground.

Here’s the million-dollar question: Can AI make the housing market more affordable—or at least less insane? No. Affordability has to do with the economy, employment, interest rates—all big, high-level issues. They will be there whether we have machine learning or not. AI could help you find a house you can afford faster and better, but unfortunately it won’t make the market itself more affordable.

This interview has been edited for length and clarity.