This elite matchmaker charges clients up to $100,000. Now, they’re building an AI

Matchmaking is an ancient dating process that stretches back thousands of years. But as online dating fatigue has begun to dominate the modern-day discourse around finding love, one company is betting that matchmaking will see a revival, and they’ve spent years developing a tool to make it happen: an AI matchmaker named Tai.

One might argue that all modern dating apps aim to serve as a kind of matchmaker; an intermediary whose purpose is to connect two singles with each other. But Adam Cohen-Aslatai, CEO of the matchmaking company Three Day Rule, says dating apps still put the onus on users to choose the right partners based on what the algorithm serves. In contrast, he explains, a traditional matchmaker uses their in-depth knowledge of the client to facilitate that process, selecting matches that aren’t just compatible on a screen, but in real life.

Now, Three Day Rule is rolling out its first-ever app designed to bring that human-centric matchmaking experience to a broader client base. The app leverages multiple AI models built upon 15 years of matchmaking data collected by the company. While Three Day Rule’s elite “white-glove” matchmaking services typically run clients around $2,000 per month, its AI matchmaker is available starting for free.

AI features have become increasingly popular on existing dating apps in recent months. But Cohen-Aslatai says none of them are fixing the larger problem: Self-serve dating apps just don’t work very well. Traditional dating apps are effective 9-11% of the time, he claims, compared with an 70-80% success rate for traditional matchmaking.

“Matchmakers aren’t self-serve; they are serving you,” Cohen-Aslatai says. “They are doing all the work for you. And they’re saying, ‘I know you better than you know yourself. I know what you need for long term relationships. And I’m not going to let you waste your time on people that are candy, I’m going to give you the full meal.’”

Is matchmaking the new swiping?

For many singles searching for love, AI-powered tools are becoming an integral part of the swiping process. Hinge now uses AI to help users craft better profile prompt responses; Tinder is set to roll out AI-powered personalized matches; the AI assistant app Rizz uses AI exclusively to help users come up with responses on dating apps; the AI relationship advice app Meeno helps men find love; and Bumble founder Whitney Wolfe Herd even announced that the company is currently working on its own AI matchmaking service. 

Cohen-Aslatai has seen it all. He previously worked at dating apps including The Meet Group, Zoosk, and Bumble’s gay dating app Chappy, and even founded and sold his own app called S’More. But, over the past few years, he’s become convinced that matchmaking is the most effective dating option out there.

He argues that, on dating apps, users are looking for a match that meets their requirements on paper—standards that, oftentimes, are preventing them from taking healthy risks. Matchmakers are trained to take those preferences into consideration while also encouraging clients to try people slightly outside their comfort zone.

Three Day Rule’s process starts by assigning three experts to each client: A matchmaker tasked with learning everything about the client’s needs and wants in a relationship, a coach to help advise clients on dates, and a recruiter who meets potential matches at places like the airport or the beach. For its highest paying VIP members, the company offers everything from personal stylists to hair and makeup services. It’s dating for people with more than just a little extra money who don’t want to leave their love lives up to chance.

Cohen-Aslatai stepped in as the CEO of Three Day Rule last October. In the past year, he says company sales have increased by a whopping 40%. Its VIP matchmaking membership—which costs a head-turning $100,000 for 16 months—has also shot up from accounting for 5% of its business to 50% in the same period. 

Last year, Three Day Rule conducted its own survey to understand how singles outside of its network are thinking about matchmaking. The report, which included 250 respondents, found that 74% would try matchmaking, while only 1% actually have. The main reason for that discrepancy is the price barrier.

“So, if you think that matchmaking works, and everyone would try it, but it’s too expensive, the question is, how do we democratize this concept?” Cohen-Aslatai says. 

How Three Day Rule built an AI matchmaker

The answer he’s landed on is Tai. Cohen-Asla”tai” and Sneha Ramachandran, general manager of Three Day Rule’s new app, have been working on perfecting this AI matchmaker for the past two years.

Whereas Three Day Rule’s services typically start at a hefty $1,500 for a three-month plan, its new app comes in a free tier, which gives users five minutes of daily voice conversation with Tai and a separate AI coach (alongside unlimited texting), as well as five ongoing conversations with matches. A second premium tier costs $99 per month and includes four guaranteed match introductions per month, 10 minutes of daily conversation with Tai and an AI coach, and unlimited conversations with matches.

Tai is built off of ChatGPT, which, Cohen-Aslatai says, has the fastest voice response rates of any other models on the market. But while ChatGPT serves as the base of the model, it’s trained on data collected by Three Day Rule. That means that all of Tai’s interactions—from their tone to the questions they ask—are informed by data collected through Three Day Rule’s human matchmakers. 

Unlike other dating services, which are primarily focused on facilitating first dates, Cohen-Aslatai explains that Three Day Rule has collected a wealth of post-date data. This information, gathered through multi-question surveys sent after every date, lets matchmakers know exactly what the client did and didn’t like about their match. All of these insights have helped to refine Tai’s matching abilities.

When users first download Three Day Rule, they’ll have to complete a mandatory photo verification process in order to join. From there, they’ll have an initial conversation with Tai (via text or voice messages), during which Tai will gather key details like important preferences, demographic and psychographic information, and relationship deal-breakers. After that discussion, Tai will begin searching the app’s database for matches. Over time, the model will make its way through 150 questions identified by Three Day Rule as important topics to understand exactly what a client is looking for.

“We’re trying to really deeply understand your personality,” Cohen-Aslatai explains. 

Part of that process means attempting to replicate the uniquely human touch of a matchmaker in AI form. If a user is giving brief “yes” or “no” answers, Ramachandran says, Tai will adjust its responses to match that cadence and probe deeper in later conversations. A chattier user will get chattier answers. Tai is even programmed to trick you by asking the same question in different ways if it suspects you may just be giving the answer it wants to hear. 

When the app identifies potential matches, both parties have to approve the decision before they’re connected online. From there, users can also chat with a second AI model in the app, which stands in for the human coach the company offers to its clients. This model has been trained using company podcasts, articles, and coaching guidelines to field questions about dates, debrief interactions, and help clients overcome personal obstacles to love. In response to a question like, “I liked him, but how can I know if he liked me?” for example, it might give three potential signals of mutual attraction. 

The app is built to offload the effort of swiping, searching, and vetting potential partners from the user onto the AI matchmaker. All the client should have to worry about, in theory, is uncovering what they really want in a partner.

Why relationship experts are wary about AI-powered dating apps

Tai is designed to make navigating the online dating world less complicated. But Treena Orchard, an associate professor at the School of Health Studies at Western University in London and author of the book Sticky, Sexy, Sad: Swipe Culture and the Darker Side of Dating, isn’t convinced.

As a researcher specializing in the intersection of sex, relationships, gender, and tech, Orchard is generally wary about the claims made by many dating apps that AI features will make dating easier—especially considering that, from her perspective, that shouldn’t necessarily be the goal.

“The problem I have with a lot of these different services is that it just really amplifies the difficulties that already exist,” Orchard says. “Dating is going to be hard no matter who’s doing it, because relationships are strange and we’re humans, and we have lots of desires and things pent up, and we’re not really good at teaching each other how to talk about these things.” 

When we start to rely on an external source to facilitate dating—whether that be a matchmaker or a digital device—she says that we become removed from the process of actually getting to know each other.

In regard to Three Day Rule’s new app, Orchard sees the utility of tapping into a network created by professional matchmakers who have experience in pulling together eligible singles. Still, she says, it’s important to recognize that, while matchmaking may be statistically more effective than dating apps, that’s also because it’s historically required a much greater financial investment. 

“Okay, it’s been trained by high end white-glove matchmakers—great—but it’s still AI,” Orchard says. “One of the critiques I have is that variety is so important. Taking a chance on someone new and someone different can open up a whole interesting Pandora’s box of experiences that you could never have calculated in your wildest dreams if you just rely on an algorithm.”

On a broader scale, she’s concerned about what it might mean to continuously find new ways to incorporate AI into our romantic relationships. For those who feel trapped in the online dating cycle, she recommends balancing the apps with trying new in-person events and hobbies—even, and especially, if it’s hard.

“The struggle is how we learn about rejection, about resilience, about failure, about one another, and how we get better at dating and become confident in ourselves,” Orchard says.

View the full article