Ditto, an AI matchmaking service for college students, has raised $9.2 million in seed funding to replace swipe-based dating apps with a weekly algorithmic pairing delivered over iMessage. The company, founded by UC Berkeley dropouts Allen Wang and Eric Liu, has attracted 150,000 signups across a few dozen colleges, and Wang says roughly 20% of matches end up going on an in-person date. Investors on the round include Gradient, Peak XV, and Scribble.
The premise is a direct rejection of the Tinder and Hinge model. There is no app to download and no photo grid to swipe through. Students text a designated iMessage number with a code, then work through an onboarding chat where an AI collects biographical basics, personality signals, dating preferences, and — for some users — photos of celebrity crushes to calibrate a type.
Every Wednesday at 7 PM, Ditto sends each user one match, one time, and one location. The user shows up. Afterward, Ditto collects feedback and feeds it back into the matching engine for the next week.
Key facts
- 01Ditto raised $9.2 million in seed funding from Gradient, Peak XV, and Scribble.
- 02The app has 150,000 signups across a few dozen colleges, with 20% of matches leading to an in-person date.
- 03Matches are delivered by AI over iMessage every Wednesday at 7 PM — one person, one time, one place.
- 04Founders Allen Wang and Eric Liu dropped out of UC Berkeley; the San Francisco company now has 12 full-time employees.
- 05Wang says 99% of investors reached out inbound, including Duolingo co-founder Severin Hacker.
Wang's pitch on the underlying algorithm is that surface-level hobby overlap is a bad signal. Ditto instead tries to infer the personality traits behind stated interests — adventurousness, individualism, aesthetic sensibility — and match on those latent variables. The company calls this predicting chemistry rather than compatibility.
It's a bet that generative AI can do a job that a decade of dating-app product design has failed at: reducing the time between matching and meeting, and cutting out the messaging layer entirely. The 20% match-to-date conversion rate is the number to watch. Tinder and Hinge do not publish comparable figures, and Wang's aside — that anyone who has used Tinder would not find 20% low — is the implicit benchmark.
Ditto's growth is currently gated by its .edu-email vetting model, which doubles as a safety mechanism. Because matches are pushed together on short notice with a specified time and place, the platform's exposure to catfishing and impersonation is higher than a traditional app where users can vet each other over days of messaging. The university-email requirement means both parties can at minimum be confirmed as students at the same school.
Scaling beyond that trust perimeter is the open question. Ditto plans to expand to more colleges and eventually beyond campuses, but every established dating platform has struggled with safety at scale. The company is based in San Francisco with 12 full-time employees and is actively hiring, including a listed role for a "Chief Yacht Officer" to run promotional events.
The marketing has leaned heavily on viral stunts — robot videos, campus takeovers, and a heavy social presence on LinkedIn and Twitter. Wang says 99% of the seed round's investor interest came inbound, and that Duolingo co-founder Severin Hacker reached out directly after seeing the company online.
The counterweight to the momentum is that dating apps are one of the most churn-heavy consumer categories in tech. Match Group's stock has traded down for years as user engagement on Tinder and Hinge has plateaued, and every new entrant that promises a fix — Bumble, Hinge itself in its early days, a long list of niche players — eventually converges on the same monetization problem: successful matches leave the platform. A 20% date rate is good for users and worse for retention.
Ditto is one of the cleaner early examples of a consumer AI product that isn't a chatbot wrapper. The AI is the matching engine and the interface — no swiping, no messaging, no profile grid. If the model works, it points at a broader consumer-AI pattern where the LLM replaces the entire product surface rather than sitting alongside it, and where a $9.2M seed at a few dozen campuses is the on-ramp to a category incumbent worth billions.
Working on something we should cover, or seeing a story we missed? Send leads, documents, or feedback to hello@aichatdaily.com. For sensitive tips, see our secure tips page for Signal and PGP options.
Spotted an error? Email hello@aichatdaily.com with the URL and the issue, or read our full corrections policy.
