U.S. startup Dextr AI announced that it has emerged from stealth after raising $6.7 million in a seed funding round led by Elevation Capital with participation from Foundation Capital, its first institutional funding. The company develops AI agents designed for hotel operations, including bookings, responding to guest requests, coordinating staff, and handling late check-ins.
Dextr was founded in San Francisco in July 2025 by Sajid Shariff and Scott Arnold. The company started with an agent called Alfred, which answers guests’ questions and assists with check-in and check-out procedures. It then added Daisy, a voice agent that handles incoming calls related to bookings, after Shariff noticed that answering callers’ questions could turn a call into an actual reservation.
From Responding to Guests to Generating Bookings
According to Shariff, one large hotel using Dextr generates between $100,000 and $300,000 in monthly bookings through Daisy. The company has also expanded into agents for group bookings and staff management, and is seeking to connect these units so that information can move from one part of hotel operations to another.
If a guest declines housekeeping service, the guest-management agent can pass that information to the operations agent to adjust staff schedules. At a property without a front-desk employee overnight, Daisy can verify a late guest’s reservation and provide room-entry instructions, in coordination with Alfred.
The company estimates that this use case could eliminate the need for an overnight front-desk shift at some properties. Shariff estimated the cost of such a shift at between $70,000 and $80,000 annually in a market such as California, while noting the difficulty of hiring workers for these periods.
What Distinguishes Dextr’s Approach?
The company says the value comes not merely from operating a single agent, but from combining multiple use cases and connecting them to hotels’ existing systems. For this reason, it sends engineers to work directly with customers, identify suitable processes for automation, and connect the agents to property-management systems such as Oracle Hospitality, OPERA Cloud, StayNtouch, Cloudbeds, and Hostaway.
Dextr also provides an interface called Doss that allows employees to direct agents via text or voice instead of learning a series of procedures inside a new software program. The company believes this approach partially addresses employee turnover, which makes training staff on complex tools and maintaining operational knowledge difficult.
Commercial Scope and Disclosed Limitations
Dextr says its agents process more than one million interactions per month across hundreds of hotels, ranging in size from properties with 21 rooms to those with 550 rooms. Customers include independent hotels and properties operating under the Hilton, Wyndham, Best Western, and IHG brands, in addition to reservation centers, lodging-rental operations, and outdoor hospitality properties such as campgrounds, recreational vehicle parks, inns, and golf courses.
The company charges a subscription based on the property and the agents used, with usage-based fees for some agents. However, it has not disclosed its revenue, profitability, or the round’s valuation. It said annual recurring revenue is growing by approximately 20% to 30% monthly, without providing detailed figures, and declined to disclose the name of the hotel brand with which it began its first direct relationship.
Why Does This Matter?
The funding marks Dextr’s transition from early experiments in guest messaging to a platform seeking to connect customer service with bookings and internal operations. This could make return measurement more closely tied to revenue and operating costs, rather than merely to the number of automated conversations. However, the model’s success will remain tied to the company’s ability to integrate its agents with different systems from one hotel to another and maintain response accuracy in situations requiring coordination among guests, staff, and reservations.
Dextr will use the funding to hire more customer-facing engineers, expand software integrations, and develop additional agents. Its team grew from three people a year ago to 21 employees, according to the article.