The next time a diner calls a busy restaurant, the voice on the other end may not belong to a person at the host stand. A growing group of AI voice hosts is now answering phones for restaurants, handling common questions and reservation tasks while human staff focus on guests in the dining room.
The trend is appearing in markets including New York City, Miami, Atlanta, San Francisco, the Bay Area, Silicon Valley, New York, and Las Vegas. The pitch is straightforward: restaurants get constant calls, many of them repetitive, and AI can pick up around the clock.
Why restaurants are using AI voice hosts
Restaurant phone calls have not disappeared in the age of Google and Resy. According to Alex Sambvani, CEO and co-founder of Slang, popular restaurants that take reservations can receive between 800 and 1,000 calls per month.
Those calls often come from last-minute bookers, tourists and visitors, older people, and people running errands while driving. Many callers ask basic questions that may already be answered online, but the phone still rings during service.
Matt Ho, the owner of Bodega SF, described the operational problem plainly: “The phones would ring constantly throughout service,” he says. “We would receive calls for basic questions that can be found on our website.”
Bodega SF became one of Maitre-D AI’s earliest clients in May. Ho also helped the founders with trial and error testing before launch. For him, the benefit was not only labor savings; it was also reducing interruptions for staff and guests.
What the systems can do
The companies in this niche offer a similar core product: an always-available phone agent for restaurants. These AI hosts can answer routine questions about dress code, cuisine, seating arrangements, and food allergy policies.
They can also help callers make, change, or cancel reservations. Some systems can transfer callers to a human employee, and some can speak multiple languages. Restaurants can buy subscription tiers that add features.
Several startups are competing for restaurant accounts. Maitre-D AI launched primarily in the Bay Area in 2024. Newo is rolling out its software at numerous Silicon Valley restaurants. RestoHost, which is one year old, is answering calls at 150 restaurants in the Atlanta metro area. Slang, which began focusing exclusively on restaurants during the COVID-19 pandemic, announced a $20 million funding round in 2023 and is gaining ground in New York and Las Vegas.
RestoHost co-founder Tomas Lopez-Saavedra said only 10 percent of calls result in a transfer to a human. That figure helps explain why restaurants see the technology as more than a novelty: if most calls are routine, an AI host can remove a large share of phone work from the dining room.
The labor pressure behind the shift
The appeal of AI phone agents is tied to the pressure restaurants face after the pandemic, including staffing problems, retention challenges, and rising labor costs. Human hosts often have to welcome guests, manage reservations, and answer the phone at the same time.
David Yang, the founder of Newo, frames the role as a practical fit for automation. He argues that restaurant phone interactions are relatively simple compared with some other service businesses, and refers to the AI as a “digital employee.”
That does not mean the technology replaces hospitality itself. It means restaurants are experimenting with which parts of hospitality can be automated without weakening the guest experience. The phone, especially for repetitive questions, is becoming one of the first places they are testing that boundary.
Where AI hosts still struggle
The technology is growing, but it is not seamless. Mathew Focht, CEO of the Emerging Fund, which specializes in restaurant tech, described the category as seeing “unbelievable, crazy growth.” At the same time, the systems still face clear limits.
Latency is one problem. Yang identified it as a pain point, and calls to AI agents showed the issue in practice: some systems asked the caller to wait while preparing an answer, while others went silent before responding.
Another issue is unpredictability when the caller moves away from the expected path. If a caller changes a reservation request or asks a vague question, AI assistants can stumble. These are the kinds of moments where a trained human host may understand intent faster than software.
Sensitive calls are another weak spot. Sambvani said humans are better suited for situations such as a diner calling to complain about food poisoning. Some restaurant operators also worry that an AI host could damage the guest experience.
Why human hospitality still matters
Brian Owens, co-owner of the New York restaurant Crave Fishbar, with three locations in the city, and the newly opened Crave Sushi Bar, tested Slang in 2021 after reopening from the pandemic and operating “as lean as possible.” The product was introduced to him by his account manager at OpenTable.
Owens initially saw the appeal. But after reviewing calls, he found unhappy customers who wanted a human employee and sometimes cursed at the chatbot. His concern was not only whether the AI could answer a question, but whether it could represent the restaurant well.
He put the difference this way: “If you’re asking a robot how the vibe at the restaurant is, versus someone who is able to explain it by not using keywords—you know the difference,” he says. “I train my host staff to answer the phone with a smile, and you’re not getting a smile when you’re using AI.”
For now, the most realistic role for an AI voice host is as a support layer rather than a complete substitute for staff. Sambvani calls it a diligent “team member.” Slang also announced an OpenTable integration on September 12, including the option to book a reservation on OpenTable by phone through the AI voice agent.
The direction is clear: AI voice hosts are becoming part of restaurant operations. The harder question is how restaurants design them so routine calls become easier without making guests feel trapped behind automation when they need a person.