Zest launches AI restaurant concierge for group chats

Zest Maps has launched an AI restaurant concierge that works through text messages, letting users get restaurant suggestions and make reservations within a conversation.

People can message the concierge directly or add it to a group chat to help choose a venue, compare options and secure a booking without switching to a separate app thread.

The product is aimed at both individual users and groups planning meals out. In one-to-one chats, the concierge suggests restaurants based on a user's preferences and past dining activity. In group messages, it weighs different tastes across participants and returns options intended to suit the whole group.

According to Zest, the system draws on dining history rather than relying only on prompts entered in a single session. Existing users can use up to two years of historical dining activity and stated preferences, while new users can start by texting the service and letting it build a profile over time.

Users who want more tailored suggestions can connect a payment card through Plaid in a browser session to share read-only dining history. Zest says the process takes less than a minute.

Reservations are handled through third-party platforms. Zest books tables through OpenTable and Resy, and provides booking links for SevenRooms, DoorDash, Tock and other services.

Mario Gomez-Hall, Founder and Chief Executive Officer of Zest, contrasted the approach with broader AI assistants that begin by asking users to describe their tastes.

"Every other AI assistant starts by asking what you like," said Mario Gomez-Hall, Founder and Chief Executive Officer of Zest. "We focus on where you've actually been. That's a completely different starting point, and you feel it in the first text."

Group planning

The concierge is designed to sit inside the conversation where dining plans are already being made, whether in group chats where friends are deciding where to go or in direct messages from users searching for neighbourhood restaurants, building lists of places they have visited or planning meals for specific occasions.

Within the same thread, it can handle discovery, planning and booking. When not in use, it remains inactive so the rest of the conversation can continue without interruption.

Behind the product is Zest's existing app and web platform, which combines a restaurant database with reviews, verified-visit data and taste profiles. The goal is to make recommendations based on observed dining behaviour rather than ratings alone.

The launch comes as AI agents are increasingly being used for consumer search and booking tasks, as companies look for ways to place those tools inside familiar communication channels rather than requiring users to open standalone apps or websites.

For Zest, the move also reflects that personal dining data can become a core asset in restaurant discovery. The company's longer-term view is that a user's taste profile should be usable across different AI services rather than confined to one product.

"Long term we think your taste profile is the product, and it should be portable to use with any agent," said Gomez-Hall. "That's what we're building toward."