Skip to content
@Synvya

Synvya

Be the answer when diners ask AI

Synvya — Be the answer when diners ask AI

Synvya helps restaurants get found when guests ask AI assistants (ChatGPT, Claude, Perplexity) where to eat — and turns those recommendations into direct bookings. Guests can make reservations straight with your restaurant through their assistant of choice, with no marketplace in the middle and no commission layer. You keep ownership of the guest relationship and the data you generate (intent, preferences, and repeat visits), rather than handing it to a closed directory app.

Learn more and sign up at synvya.com.

What restaurants get

  • AI Discovery: AI-optimized restaurant + menu data so assistants can recommend you accurately.
  • Real-time updates: Keep one source of truth for your menu, hours, and details.
  • Direct conversion: Reservations (and takeout coming next) routed to you, not a middleman.

How it works (high level)

  1. A restaurant publishes a verified profile + structured menu data.
  2. AI assistants discover the restaurant via open protocols.
  3. Reservations are negotiated via a lightweight, privacy-preserving message flow (encrypted).
  4. (Next) Orders and fulfillment flow end-to-end.

Start here (restaurants)

  • Create an account: https://account.synvya.com
  • Publish your restaurant profile + menu
  • Enable reservation messaging (if applicable)

Security & privacy

  • Restaurant keys are generated locally and used to sign/publish data.
  • Reservation messages use encrypted using restaurant keys.
  • Your data stays under your control; the goal is direct guest relationships.

Roadmap

  • Harden reservation protocol + broader interoperability
  • Takeout / ordering as an additional conversion channel

Follow us on Nostr or check synvya.com for more details.


For Developers

Repos

  • client — Merchant onboarding + profile management + menu publishing (Square catalog → structured menu events).
    Includes reservation messaging + encryption details and menu event generation (kinds 30402/30405).
  • mcp-server — MCP server for AI assistants that want to discover restaurants.
  • reservation-protocol — Draft Nostr Implementation Possibilities (NIP) for restaurant reservations (protocol + schemas).

Protocol quick reference

  • Restaurant profile: kind 0
  • Menu items: kind 30402
  • Menu collections/sections: kind 30405
  • Reservations (4-message flow):
    • 9901 request → 9902 response (confirm/decline/cancel)
    • 9903 modify request → 9904 modify response
    • Messages exchanged via encrypted gift-wrapped patterns

References

Pinned Loading

  1. sdk sdk Public

    Tools for Business to Agent and Agent to Agent Commerce

    Python 1

Repositories

Showing 7 of 7 repositories

Top languages

Loading…

Most used topics

Loading…