How it works

Most apps show what’s popular.
Srvn learns what you’d actually like.

An AI dining agent built on a personal taste graph that compounds with every meal.

The State of Restaurant Choice

Restaurant choice is still broken.

The issue is not a lack of places. It is scattered, low-trust, non-personal information.

Problem 01
Anonymous ratings

Generic scores don't explain whether a place fits your taste.

Problem 02
TikTok / Instagram

Great for inspiration, weak at turning desire into a dinner decision.

Problem 03
Google nearby

Shows what's close or popular, not what this person or group should choose.

Problem 04
Group chats

Endless back-and-forth with no memory of the last meal.

The problem is not discovery. It’s decision-making.

How the Agent Decides

It reads the full picture.

Personal taste signals+friend & group context+restaurant and dish data+occasion, budget, location, vibe+past decisions

The Moat

The first real taste graph for food.

The moat isn’t the interface. It’s the compounding memory created from repeated dining decisions.

TasteGraphDish notesPhotosReviewsRatingsSavesFriendsOccasion + contextWhat was chosenGroup eventsWho suggestedAttendance

Join the waitlist.

Get early access to the first AI dining agent that learns your taste. We’re rolling out invites in waves.

Get early access