TROY AI · PRODUCT DESIGN
TROY AI · PRODUCT DESIGN
TROY AI
Designing an LLM-powered reservation system that helps hospitality teams turn every call into a confirmed guest.
Designing an LLM-powered reservation system that helps hospitality teams turn every call into a confirmed guest.
Troy AI is a reservation ecosystem for hospitality teams. As Founding Product Designer, I led the product from early research to an AI phone-answering host, an iPad management system, and a unified web booking experience.
Troy AI is a reservation ecosystem for hospitality teams. As Founding Product Designer, I led the product from early research to an AI phone-answering host, an iPad management system, and a unified web booking experience.
ROLE
Product Designer
SCOPE
UX strategy · workflows · visual system
YEAR
2024
THE CHALLENGE
Restaurant teams were missing valuable calls while juggling bookings across disconnected tools. The product needed to make AI-assisted reservations feel dependable, understandable, and effortless in the pace of service.
Restaurant teams were missing valuable calls while juggling bookings across disconnected tools. The product needed to make AI-assisted reservations feel dependable, understandable, and effortless in the pace of service.

Signal over noise
A clear operational hierarchy surfaces today’s reservations, call outcomes, and exceptions without requiring teams to hunt through scattered systems.
Signal over noise
A clear operational hierarchy surfaces today’s reservations, call outcomes, and exceptions without requiring teams to hunt through scattered systems.
Shared context
Each reservation keeps the conversation, booking history, and follow-up action connected—so the AI feels like a trustworthy teammate, not a black box.
Shared context
Each reservation keeps the conversation, booking history, and follow-up action connected—so the AI feels like a trustworthy teammate, not a black box.
OUTCOME
Troy AI processed over 13,000 calls, recovered more than 1,000 manual hours, and helped six Bay Area locations protect over $50,000 in revenue.
Read full case study