Nandini Mullaji grew up close to matchmaking; her grandmother practiced it, and Mullaji later worked on dating and education projects, including Bumble's launch in India. What she believed conventional dating apps lacked was not the number of profiles but the depth of information a good matchmaker collects before introducing two people. A handful of prompts and a history of swipes could not replicate the reasons she would pair one person with another.
A matchmaker's knowledge turned into a product
Mullaji co-founded Sitch with Chad DePue, a technology executive who had been chief technology officer at Whisper and later worked at Snap. The pair used Mullaji's framework—more than 75 matching parameters—as the foundation for an AI-assisted service. Sitch launched in New York in December 2024, positioning itself as a setup product rather than another feed of faces.
Making onboarding intentionally slower
A new member answers close to 50 questions by text or voice. The conversational intake is meant to surface values, patterns and preferences that are difficult to infer from photographs. The system proposes a person and explains points of compatibility and contrast. If both members accept, it opens a group chat, and it can collect feedback after a real-world date to refine future introductions.
The payment model changes the incentive. Instead of charging a recurring subscription for continued access to the pool, Sitch sells packs of setups. The company therefore earns when it makes an introduction, not when a user keeps scrolling. It is still a commercial dating product, and an introduction cannot guarantee chemistry, but the unit of value is closer to what the customer came to accomplish.
Trust is both the data and the moat
In June 2025, TechCrunch reported a $5 million seed round led by M13 with participation from a16z speedrun, bringing total funding to $7 million. The company was operating thousands of introductions in New York and preparing to expand into San Francisco and Los Angeles. City-by-city growth gives Sitch density before breadth, a useful discipline for a network whose results depend on having enough compatible people nearby.
The model also asks users to disclose unusually personal information. Sitch has said people share openly because the intake is not posted as a public profile, but that trust has to be renewed through privacy choices, honest explanations of AI and careful claims about outcomes. Its marketing challenge is not merely persuading singles to download another app. It is showing why a slower, more revealing process deserves a place in their search for a relationship.
