Two People Run a 6,000-Member Club. Pranay Periwal Built the AI That Makes It Possible.

Tova has 6,000 vetted members, has run more than 450 dinners, events, and trips in two years, and operates in two cities without a venue. It is run by two people. Pranay Periwal, who founded the club with Athena Au in 2024, estimates that it now takes him between five and ten hours a week. Much of the work that used to fill his weeks is now done by AI agents he built himself, one task at a time, while the club was growing around them.
"For the first fifty or sixty events, Athena and I did everything," Periwal says. "We read every application, chose every table, hosted the evening and wrote the follow-ups afterwards. We needed to know what a good night looked like before we could ask software to produce one."
Matching came first. Which ten people end up in the same room is the decision the club depends on, and Periwal and Au worked out how to make it early, combining psychological compatibility theory with AI. "Matching was the first thing we had to get right, because it is what the club is," he says. "The model works from a survey every member fills in: communication style, personality type, interests, cultural background, how they handle disagreement."
With matching in place, he automated the parts of the job that had an obvious correct outcome. Application processing came first: reading a submission, checking it against what the club looks for, and sorting it. Application feedback followed, the messages that tell an applicant where they stand. Then came event discovery, finding the formats and occasions worth running, and host discovery, finding the venues and event organisers to run them with.
Next, he turned to internal tooling: a system that lets the team retrieve anything about a member, an event, or a host without searching through records, and an onboarding AI for new hosts. The system can take a venue or organizer working with Tova for the first time through how the club runs an evening without requiring either founder to be on the call. That reduced another recurring operational task while keeping the process consistent as Tova added more events and partners.
Some of the work stayed with people. Application evaluation, as distinct from processing, still has a person in the loop. "Reading an application is easy to automate. Deciding whether someone belongs is harder," Periwal says. "The system does the sorting and flags what it is unsure about, and a human makes the call." On the day of an event, he says, no software replaces the host running the room, and the coordination with venues and organizers in the run-up still needs a person at the other end of the conversation.
The effect is a club that can add a city without adding staff at the same rate. Tova is currently trialing chapter heads in Hong Kong and Singapore, members who run a local calendar using the same tools. London is the next city, with a launch planned for the first quarter of 2027. It will be Tova's first market outside Asia, and it will open the way Hong Kong and Singapore did: no clubhouse, a vetted membership, and dinners and events assembled by the matching system. The chapter-head structure is how a two-person team intends to run it, with New York and San Francisco planned to follow.

Periwal built the system in-house with some off-the-shelf tools connected to it, and he declines to name them. The operating model also reflects clear limits on what the software handles. The technology takes on repeatable work, while people remain responsible for the decisions and interactions where context matters more.
Since June 2025, Periwal has worked at Color, a San Francisco startup backed by NFX and a16z speedrun. He is its founding product manager and has been with the company through its first fundraise. The company describes its product as an AI guardian angel for everyday life, with a face and a voice, working across the tools people already use, including email, calendars, messages, and notes.
Periwal sees a connection between the two projects. "Tova uses software to get the right people into a room, and then it gets out of the way. Color does the same for everything else," he says. "She looks after your whole life, from work and money to family and travel, and handles things before you've noticed they need doing. AI moves so fast that most people can't keep up. She keeps up for you. You have someone to rely on, and you get to be present for the people and things you care about. That's what we mean by making life more colorful."
At Tova, that has meant five to ten hours a week instead of a full-time job. At Color, he wants to give that time back to everyone who uses it.
© Copyright IBTimes 2026. All rights reserved.

