Dextr AI, an artificial-intelligence platform building an agentic workforce for the hospitality industry, has raised $6.7 million in seed funding in a round led by Elevation Capital, with participation from Foundation Capital. Announced in late September 2026 as the company emerged from stealth, the round marks Dextr AI's first institutional financing.
Founded in July 2025 by Sajid Shariff (CEO) and Scott Arnold (CTO, formerly at Meta), Dextr moves past generic chatbots with a fleet of domain-specific AI agents. Daisy is a voice-based reservation agent handling bookings, modifications and upselling in more than 90 languages, processing tens of millions of dollars in phone reservations each month. Alfred manages guest experience and check-ins, Manny automates back-office operations like roster scheduling, and Cosmo drives direct bookings and digital marketing. Alongside the raise, the company unveiled Doss Gen 1, a command layer that lets hotel staff direct the AI workforce in plain language via text or voice.
The platform is already live at scale: Dextr says it powers more than one million monthly interactions across hundreds of properties in the US, Canada, the UK and Europe, including franchised properties operating under the Hilton, Wyndham, Best Western and IHG brands. One reservation centre reported $80,000 in incremental revenue within 60 days of deployment, while a destination resort cut operating costs for $120,000 in annual savings. The company is SOC 2 Type 2, PCI and GDPR compliant.
The fresh capital will be used to expand its AI research and product engineering teams, deepen integrations with hospitality software, and grow across global hospitality markets — including India. That India angle is deliberate: Dextr's analysis of Indian contact centres found a 42% missed-call rate, and in hospitality, a missed call often means the booking flows to an online travel agency or a rival property.
Why it matters for founders
Dextr AI shows how vertical AI agents are moving from demos to P&L impact — its customers measure the product in incremental revenue and labour hours, not pilot counts. If you're building AI for a service industry, the lesson is to own a workflow end-to-end (payments included) and land with measurable ROI numbers, because that is what turns a pilot into a deployment — and a seed round into a breakout one.