Screenshots
See it in action.
⤢
⤢
⤢
⤢
⤢
A multi-tenant SaaS platform delivering WhatsApp-based AI concierge services for hotels. Four microservices, RAG-powered intelligence, real-time dashboards, and 86 API endpoints — built from zero to production in 8 months.
Hotels face constant staff overhead from repetitive guest questions, language barriers with international visitors, revenue leakage from guests unaware of available services, and inconsistent guest satisfaction across shifts and seasons.
An AI concierge that lives where guests already are — WhatsApp. No app downloads, no accounts. Guests text naturally and get instant, contextual answers about the hotel, local attractions, weather, restaurants, and more — all powered by RAG over hotel-specific documents.
Guardian filter classifies every message before it reaches the main agent, preventing prompt injection and off-topic abuse.
Per-tenant vector stores built from uploaded PDFs, DOCX, and Markdown. Guests get answers grounded in actual hotel information.
8 agent tools — weather, ski conditions, local businesses, tickets, room info — loaded dynamically per tenant configuration.
Staff monitor all conversations live with sentiment analysis, word cloud analytics, and one-click human escalation.
Complete data isolation between hotels. Each tenant gets its own vector store, knowledge base, tool configuration, and branding.
AI creates tickets for complex requests and escalates to human staff with full conversation context preserved.