Zest launched an AI concierge that recommends restaurants, runs group votes and books tables by text, without a separate app.
Zest has launched an AI restaurant concierge that finds places to eat and books tables by text. The restaurant discovery startup introduced the agent on Thursday, TechCrunch reported. Users can message it directly at a dedicated phone number or add it to a group chat. This article explains how it works, what data it uses and where it fits among rival text agents.
AI Restaurant Concierge Works Over Text
Zest’s agent lives in a text thread, so users need no separate app. People can ask for recommendations by cuisine, craving, occasion, dietary needs or neighborhood. They can also ask how busy a place is, whether it has Wi-Fi, or what is on the menu. Once they choose, the agent books the table inside the same conversation.
TechCrunch tested the tool and found its suggestions fairly good. The picks matched the time of day, party size and location. The reporter also found it convenient to see open times without leaving the chat. TechCrunch compared it with other text-based agents, such as Instinct and Poke.
Group Chat Dining Gets a Voting Tool
The agent can also join a group chat. It blends the tastes of everyone in the thread. When friends disagree, an in-thread vote helps settle where to eat. Then the agent books the table without anyone leaving the conversation.
Restaurant Booking Partners and Pricing
TechCrunch reported that the agent checks availability and books through OpenTable, Resy and SevenRooms [UNVERIFIED — SOURCE NEEDED]. A second outlet that covered the launch, The Tech Buzz, did not name any booking partners. Booking and ordering are Zest’s first planned revenue streams, and food ordering is coming next. The coverage reviewed did not disclose pricing.
Zest began testing the concierge a month earlier, TechCrunch said. The company held the launch until booking worked. The company may later expand beyond restaurants into other recommendations, such as shopping. For now, dining remains the focus.
Plaid Data Powers Personalized Dining Picks
Zest was founded in November 2024 and launched its restaurant app earlier this year. Users link a credit card through Plaid, a financial data service used by many banks and budgeting apps. As TechCrunch reported at launch, Zest keeps only food and drink purchases and discards the rest. Those purchases build personal dining maps that drive the recommendations. Users can also follow friends and creator-curated profiles. The app draws on more than 80 million web reviews as well. At launch, TechCrunch said a notes feature and a Fresh Picks list for new restaurants were coming.
The startup raised $1.8 million in pre-seed funding from Alexis Ohanian of 776 and Steve Jang of Kindred Ventures. Founder Mario Gomez-Hall described the mission in a May press release. That release described card linking as optional and said the iOS app comes from San Francisco. It also said Zest draws on social content from TikTok, Instagram and Reddit. Users can see a live feed of friends’ dining and ratings, too.
“Our vision has always been to rethink how people discover food…” — Mario Gomez-Hall, CEO and Co-Founder, Zest Maps
Rival AI Agents in Text Messages
Zest joins a growing group of agents that people can text like contacts, TechCrunch’s roundup shows. Poke launched in March as a general assistant for calendars, planning and smart-home controls. In June, it became the first AI agent that Apple approved for Messages for Business. Instinct, still in private beta, raised $1 billion in September at a $10 billion valuation. Early users have used it to plan trips and cancel subscriptions.
Zest’s pitch is narrower than those tools. It focuses on dining and builds its picks from where users actually eat. Other agents are moving into money and teamwork instead. For example, Binance’s AI agent trading tools let bots trade from separate sub-accounts. Meanwhile, OpenClaw 2.0 adds shared workspaces where teams and agents can work together.
Ultimately, Zest is betting that one text thread can replace several restaurant apps. Its tool combines recommendations, group decisions and booking in one place. The company still has to show how it will earn money and which partners it relies on. If it succeeds, the agent could grow well beyond restaurants.








