It was during a walk with their kids when neighbors Matt Schwartz and Dan Zhang came up with the idea for Ramsi, a platform they cofounded that leverages agentic AI to help hotel owners and operators maximize room revenue.
Zhang, associate dean for research and academics at Leeds School of Business at CU Boulder, is a veteran of pricing optimization, with experience spanning airlines, ride sharing and large-scale e-commerce.
That experience sparked a bigger idea: bring advanced, AI-driven pricing intelligence into hotel revenue management.
Schwartz, who is also chief technology officer at Sage Hospitality Group, recalled, “One day, he said, ‘I have all this amazing math in my head, more sophisticated than anything going on in the hotel industry.”
From there, the founders focused on revenue management as one of the most data-heavy and variable-driven functions in a hotel—and began building what Schwartz described as a “fully autonomous, AI-based pricing service.”
A third cofounder, David Li, a professor at the City University of Hong Kong, joined after working with Zhang at Alibaba, while Jeff Loucks came aboard as the VP of revenue strategy to translate advanced math into practical hotel workflows.
Ramsi’s approach is rooted in agentic AI—a concept that is quickly becoming one of the industry’s most buzzed-about terms, but one that the Ramsi team says delivers tangible benefits when applied to revenue management.
“Agentic AI refers to a system of specialized agents,” Loucks explained. “Each one is responsible for a specific task. They all work together towards the broader objective, which is to create those pricing recommendations.”
Instead of relying on a single model or static rules, the system distributes decision-making across multiple demand inputs. “Each of our agents focuses on a defined problem or demand factor, like competitor pricing, events, weather, historical data or recent pickup,” Loucks said.
From the hotel’s perspective, Ramsi is built for flexibility—supporting both revenue-managed properties and operations where pricing is still handled by a GM or rooms leader.
“Our system is capable of fully updating rates throughout the day,” Loucks said. “If a hotel prefers just to review the recommendations, they can do that, too. They can automate and override.”
He noted that the platform continuously evaluates “tons of demand signals—competitor rates, booking pace, events, weather—to create those price recommendations and then update them automatically.”
Ramsi has also leaned into simplicity on the operational side—particularly in onboarding and usability. “We provide all of this through a very simple interface,” Loucks said. “We get through the training process in about 30 minutes, and we manage the entire onboarding process internally.”
Today, Ramsi is live in about 75 hotels across North America, Europe and Asia, with customers ranging from boutiques to larger independent resorts. “We work with a range of hotel types, from small, 10-room properties up to resorts with more than 300 rooms,” Loucks said.
Ramsi is also developing a free tool that will allow hotels to go to the company’s website to receive a complimentary pricing analysis by entering basic property information.
Additionally, Schwartz noted, the company’s PhD-level data scientists are developing a foundational model—an AI model trained on massive datasets that develops broad general knowledge—for pricing.
“It’s being introduced into our system now and will be fully introduced later this year,” he said. “That will make the pricing capability significantly more advanced.”
The company was also recently invited into the MIT Startup Exchange, which, he said, “opens up access to MIT-affiliated companies and their global alumni base. We think that’s a unique advantage as well.”
