222 Wants AI to Walk Three Strangers Into the Same New York Bar

The Y Combinator-backed app raised $2.5M to match groups for offline meetups, betting questionnaires beat swipes.

About 222

Published

On the 222 platform, a Friday night begins with a questionnaire of nearly 100 prompts, runs through a machine-learning matcher, and ends with a group of strangers sitting down together at a curated venue in New York. That is the wager Danial Hashemi, Keyan Kazemian, and Arman Roshannai have placed: that the bottleneck on adult friendship is not discovery, it is orchestration, and that software should do the orchestrating.

The company, founded in 2023 and now based in New York after relocating from Los Angeles, raised $2.5 million in seed funding in February 2024 [Business Insider, Feb 2024]. It carries 16 employees per its Y Combinator profile [Y Combinator]. Backers include First Round Capital, Canaan, Alumni Ventures, HighSage Ventures, Flintlock Capital, Gaingels, and, less typically for a consumer social app, Progressive Insurance.

The bet

222 sells a product that older social networks largely abandoned: the in-person meetup, scheduled for you. Users complete a long intake covering geography, interests, and personal views (Business Insider notes the questionnaire reaches into territory as personal as opinions on abortion), and the system clusters them into small groups for a specific time and place [Business Insider, Feb 2024]. The company describes its work as engineering serendipity, matching strangers for bespoke real-life experiences at hyperlocal venues [LinkedIn].

The wedge is the curation layer. Eventbrite sells you a ticket. Bumble hands you a profile. 222 books a table for four people who, on paper, should get along, then nudges them to show up. It is a B2C product priced and packaged around the night out itself, available through an iOS app launched in 2024 [Apple App Store] and an RSVP flow at rsvp.222.place.

Why it could be big

The market context is unusually friendly. Business Insider, in a May 2024 survey of the category, framed the moment as a loneliness epidemic giving rise to a new crop of startups aiming to help people connect in real life [Business Insider, May 2024]. That same piece grouped 222 with peers including Timeleft, Pie, Saturday, and The Breakfast, all chasing variants of the same insight: that post-pandemic adults want fewer feeds and more dinners.

The investor roster suggests the thesis is being underwritten seriously. First Round Capital and Canaan are not casual consumer-social tourists. The presence of Progressive Insurance is the more interesting tell: a strategic check from a Fortune 100 carrier into an IRL meetup app implies someone inside Progressive sees a longer-term link between offline community formation and the kind of behavioral data insurers care about. Y Combinator backed the company in its earlier seed round in April 2023 [Y Combinator].

Metric Value
Seed round, Feb 2024 2.5 $M
Employees 16 people

The team and traction

Hashemi, Kazemian, and Roshannai built the original product in Los Angeles before moving the company east in early 2024, a relocation Business Insider tied directly to the seed raise and the iOS launch [Business Insider, Feb 2024]. Roshannai's public title, "Engineering Serendipity @ 222" [LinkedIn], is on-brand to the point of being a product spec. The headcount of 16 [Y Combinator] is small enough that the team is still operating as a single product cell, which matters for a company whose core challenge is iterating the matching algorithm against live event outcomes.

A March 2024 Business Insider piece on how new social-networking startups are approaching monetization placed 222 among a cohort experimenting with models that look less like Instagram's ad stack and more like ticketed or membership economics [Business Insider, Mar 2024]. The company's own RSVP funnel suggests the same: revenue, when it scales, is likely to come from the night itself rather than from selling attention back to advertisers.

The honest counterfactual

The bear case on IRL social is well-rehearsed. The category is crowded, with Timeleft, Pie, Saturday, and The Breakfast all chasing overlapping demographics, and the unit economics of curated offline events are harder than software margins suggest: every dinner is a logistics problem, and no-show rates can collapse the experience for the people who do turn up. Business Insider's May 2024 survey of the space flagged exactly this scaling tension across the cohort [Business Insider, May 2024].

The bull answer, and the reason 222's specific approach is interesting, is that the questionnaire-plus-ML matching layer is a defensible asset in a way that a generic Meetup-style listings page is not. If the model genuinely produces groups that click, retention compounds in a category where most competitors rely on novelty. The company's framing, repeated across coverage from dot.LA and Global Dating Insights, is that the matching quality, not the event inventory, is the moat [dot.LA] [Global Dating Insights].

What to watch

The next 12 months should answer two questions. First, geography: 222 is a New York product today, and the cost structure of city-by-city expansion will determine whether the seed round buys two markets or ten. Second, the round after this one. A $2.5 million seed in February 2024 from a syndicate that includes First Round and Canaan implies a Series A conversation is already on the calendar; the metric those investors will price off is almost certainly repeat attendance per user, not downloads. A DNYUZ feature published in February 2026 framed the company's evolution toward AI-mediated romantic introductions as well [DNYUZ, Feb 2026], hinting that the product roadmap is widening from platonic group dinners into dating-adjacent territory, which is a larger TAM and a more contested one.

The interesting question for readers: if an algorithm can reliably seat four strangers at a table in Brooklyn who actually want to see each other again, is that a social network, a dating app, or something the category does not yet have a name for, and which incumbent loses first?

Read on Startuply.vc