Dextr AI has raised $6.7 million in seed funding led by Elevation Capital, with participation from Foundation Capital, to sell hotels something more specific than AI for hospitality. The company, co-founded last year by Sajid Shariff and Scott Arnold, builds AI agents that don’t just answer guest questions - they complete reservations, process payments over the phone, and run parts of a property’s sales and marketing, already live across independent hotels and franchised properties flying the Hilton, Wyndham, Best Western, and IHG flags across the US, UK, Canada, and Europe. India is one of the markets it’s expanding into next.
Answering a question and closing a transaction are different products
Most guest-facing AI in hospitality to date has been a chatbot bolted onto a booking widget - useful for answering when checkout is, useless for actually booking the room. Dextr’s agents handle voice calls end to end, including taking payment, which is the harder engineering problem and the more defensible pitch: a property doesn’t need another FAQ tool, it needs the phone answered and the reservation closed the way a trained front-desk agent would do it. The company says it’s now processing tens of millions of dollars in phone reservations every month.
The vertical focus is doing more work than the AI
A general-purpose AI agent has to win the argument that it’s better than a human at everything. A hospitality-specific one only has to win it at the handful of workflows - reservations, guest requests, staff scheduling - that a mid-size property actually struggles to staff at 2 a.m. or during a surge in call volume. Narrowing the pitch to one industry, rather than claiming to replace an entire category of labor, is what makes the sale easier to close with an operator who has seen plenty of software promise more than it delivered.
What Dextr is really selling operators is coverage, not novelty
No hotel general manager wakes up wanting an AI agent. What they want is a phone that gets answered at 3 a.m. without hiring a night shift for a property that doesn’t do enough volume to justify one. Framing the product around coverage the property can’t otherwise afford, rather than the novelty of the technology, is why an operator running a franchise under a global brand’s strict service standards would even consider handing reservations to software in the first place.
The interesting bet isn’t that AI can talk to hotel guests. Plenty of tools already do that badly. It’s that a narrow, transactional promise - answer the phone, close the booking, get paid - beats a broad one every time an operator is the one deciding whether to trust it.