HotelOS
Multi-tenant operating system for hotel chains: RAG knowledge base, checklists, AI scheduling and dashboards.

Overview
HotelOS is a multi-tenant operating system for hotel chains, born from a simple observation: in hotels, knowledge exists, but it is scattered across manuals, files, emails and people's memory. Each department works well, but each with its own version of the truth.
The platform brings together the corporate knowledge base, a role-filtered RAG chatbot, operational checklists, AI-assisted shift planning and the managerial dashboard. The hierarchy is explicit — platform, tenant, hotel, department, user — because in hospitality boundaries matter.
The chatbot answers on hotel SOPs filtered by role: a receptionist does not see housekeeping procedures and vice versa. The knowledge base is centralized, with versioning and AI translation, so an updated procedure is updated everywhere.
Checklists support photo evidence and sub-tasks, while scheduling is driven by a conversational Shift Copilot: you describe the need and the system proposes the shifts. Everything is built with Next.js, React, TypeScript, Supabase with pgvector, OpenAI and the Vercel AI SDK.
The numbers tell the value: 117% ROI in year one, sharp reductions in information-search time and shift planning. In a real case monthly planning dropped from 6-8 hours to 30-45 minutes — the difference between a lost afternoon and a quarter of an hour.
Results
The value proposition measures 117% ROI in year one, with sharp reductions in information search, shift planning and unfinished tasks; in a real case monthly planning dropped from 6-8 hours to 30-45 minutes.
What's inside
- Role-filtered RAG chatbot over hotel SOPs
- Centralized knowledge base with versioning and AI translation
- Checklists with photo evidence and sub-tasks
- AI shift scheduling with a conversational Shift Copilot