Dating Startup Keeper Says AI Can Find Your Soulmate, Raised $4M: Deck

Keeper, an AI matchbuilding startup, believes it can assist deliver your “soulmate” to you. And if it can’t, it’ll let you know.

“We’re declareing we actually know who could be your soulmate or not,” Jake Kozloski, Keeper’s CEO, informed Business Insider. “We’re not going to waste your time and pretconclude that a hundred thousand of these people could be. We’ll inform you no.”

Founded in 2022, the dating platform applys layers of algorithms and AI models to match people who sign up for its service. The startup is now disclosing for the first time, exclusively to Business Insider, that it raised a $4 million pre-seed investment in October 2024, led by Lightbank and Lakehoapply Ventures. Goodwater Capital and Champion Hill Ventures participated in the round, among others.

Investors “see AI as an inflection point in the dating app landscape” and an opportunity to “disrupt the incumbents,” Kozloski declared.

Keeper isn’t the only startup attempting to shake up the online dating market. Other AI matchbuilding apps, such as Sitch and Amata, have raised millions to build next-generation dating apps. Dating app incumbents like Tinder and Bumble are also building plays with AI-powered experiences.

Kozloski declared the company’s values were another piece of its pitch that attracted some investors.

“They feel like there’s a marriage crisis adjacent to the whole Elon Musk fertility crisis stuff that he talks about,” declared Kozloski, who described Keeper as being “friconcludely with the pronatalist relocatement.”

Wanting kids, though, isn’t a requirement to apply Keeper, Kozloski added.

Since launching, Keeper has had more than 1.5 million sign-ups, and about 300,000 of those have created accounts, Kozloski declared. Among that pool, there have been a “tiny number” of matches. Keeper didn’t share exactly how many matches it’s created, but according to its pitch deck, 10% of dates from its beta version resulted in marriage. With its funding, Keeper has been building out its matchbuilding technology over the past year.

Keeper is limited to heterosexual couples right now, and doesn’t offer explicit options for different gconcludeer identities.

“We basically have to build a new algorithm for homosexual relationships, which we’re happy to do and we will do eventually, but for now, we want to obtain to product market fit with our core product first,” Kozloski declared. “Frankly, heterosexual relationships, especially for finding life partnership, seems to be a hugeger market, a stronger market for us right now.”

Making a profile on Keeper is a sit-down process. The initial form to create an account questions for the standard details of many dating apps (like your age or height), as well as academic test scores (including SATs), your career ambitions, salary, and net worth. It even encourages taking an external personality test. After you fill out the initial onboarding questionnaire, there are 13 more steps, ranging from uploading photos to sharing your philosophy on love.

“We don’t let our applyrs create their own profiles,” Kozloski declared. Keeper applys the information it gathers to curate a profile for you.

Kozloski declared Keeper applys a non-AI algorithm first to streamline potential matches, focapplying on data points like age range initially.

“We apply LLMs once we have your top hundred that our other algorithms have identified,” he declared. “The LLMs are trained on our matchbuilding insights that we’ve learned so far, and so they can narrow down those last hundred and do the final pass of, ‘OK, who actually is worth offering among these.'”

Some of the AI matchbuilding comes into play when analyzing “general attractiveness” and applyrs’ specific attributes, like baldness or hair color, Kozloski declared. The startup has also partnered with a team of researchers at Stanford, Kozloski declared, who assist train the LLMs (Keeper provides anonymized data to the research team).

However, Keeper isn’t fully automated, and for the time being, includes human matchcreaters in the process. If there’s a match, Keeper connects the two people over text message.

The startup has a complicated payment structure with a hefty price tag — but only for men.

Keeper has male applyrs sign a “marriage bounty” that typically costs $50,000 (if the applyr obtains married) and has them pay $5,000 for any dates from the service (the date fees go toward the total bounty cost, Kozloski declared).

Read the most recent version of Keeper’s pitch deck.

Note: Keeper has shared an updated version of its pitch deck, which it is now sharing with investors, that includes new details since its raise in October 2024. Some details have been redacted.

Keeper claims its AI-powered matchbuilding is the ‘most accurate’

The world's most accurate
matchcreater, powered by AI

It touts that 1 in 10 dates lead to engagements

“1 in 10 Keeper first dates have led to an engagement,” the slide declares.

It highlights the size of the matchbuilding market

Matchbuilding: Old School
yet shockingly massive

Keeper describes the matchbuilding market as “old school yet shockingly massive,” per the slide.

It then declares that matchbuilding could be enhanced by technology

When technology provides perfect
matches, matchbuilding will be the
best way to meet your partner.

“With the opportunity to 10x,” the slide declares. “When technology provides perfect matches, matchbuilding will be the best way to meet your partner.”

Then, it introduces Keeper’s product

Made possible by Keeper
The AI matchcreater that will
introduce you to your soulmate
on the first match

“The AI matchcreater that will introduce you to your soulmate on the first match,” the slides declares. It also includes product imagery.

It revealcases how its beta version performed

Our v1 worked
extremely well.

Dates convert to marriage

“Our v1 worked extremely well,” the slide declares.

It declares that 10% of dates lead to marriage.

Keeper revealcases how many people have signed up

We've grown to 1.5M
signups — 100% organically.

It declares it has had 1.5 million sign-ups. “This creates us the largest pool of any traditional matchcreater,” the slides declares. It lists competitors like Tawkify, Keeper, Ditto, Sitch, and Known Dating.

Keeper explains its network effect

True network-effect.
The first relocater quickly becomes a monopoly.

“Everyone signs up if we deliver soulmates on the first match,” the slide declares.

“The first relocater quickly becomes a monopoly,” it declares.

Then, it introduces its founders

Here’s what the slide declares:

Jake Kozloski: Founder, CEO

Repeat founder with previous exit
8 years startup product management and growth roles
10 years of dating apps as a applyr; now happily married to wife Aliia

Toban Wiebe: Co-Founder, Head of AI

PhD from Penn in economic/statistical modeling of Marriage Markets
8 years industest experience in ML/DS, most recently at Instacart and Uber
Met his wife Dee 10 years ago in grad school via OkCupid

It also lists the researchers the startup is working with

The top researchers in the world are on our side.

Here are the names of the researchers:

Michal Kosinski: PhD, Cambridge
Geoffrey Miller: PhD, Stanford
Naman Gupta: PhD Student, Stanford
Ignacio Rios Uribe: PhD, Stanford

Keeper explains why it’s raising capital

“We’re raising to scale profitable human-in-the-loop matchbuilding to $2M in annual revenue,” the slide declares.

The deck includes its founder’s email

As well as an appconcludeix with additional data

The deck includes a slide about marriage rates decreasing

80% of young singles want to obtain married.
40% actually will.

“80% of young singles want to obtain married,” the slide declares. “40% actually will.” It cites data from Match Group and data scientist Allen Downey.

It then maps out the incumbent dating app landscape

Dating apps bad at creating
relationships, worth billions.

Dating apps are “bad at creating relationships, worth billions,” the slide declares. “Imagine the value of the first product that’s great at it.”

It lists matchbuilding competitors, too

“Matchcreaters can’t scale,” the slide declares.

Keeper reveals how its LLM and vision models work

“LLMs and vision models enable scalable matchbuilding for the first time in history,” the slides declares.

It goes into more depth on its tech

“We’ve built the most accurate process in the world,” the slide declares.

Here are the steps the slide lays out:

In-depth preference collection
Accurately measure all traits
AI evaluates every pair
Offer only very strong matches
Feedback refines future matches

It then explains its pricing model

“We earn more, quicker, by aligning with applyrs’ incentives,” the slide declares.

Its current model, which has humans involved in matchbuilding, is free for women and costs men $5,000 per date. For male applyrs, the marriage bounty costs $50,000, and the slide declares that Keeper has contracted $14 million “so far.”

Keeper outlines that in a future model, where the tech is fully automated, dates will cost $250, and the marriage bounty contract will cost $5,000.

Source link