hotel-ai

HAIP (Hotel AI Platform) — Open-source, API-first hotel PMS. TypeScript/Node.js/NestJS.

Claude AI Plugin Marketplace — Enterprise-grade hotel operations intelligence suite. 10 skills for revenue management, forecasting, P&L analysis, labor optimization, reputation intelligence, and more.

Hotel AI Multi-Agent System – 五星级酒店智能助手，6 Agent 协作系统

🏨 Hotel AI Skills 2026: Enterprise Revenue & P&L Optimization Suite

A natural-language analyst over your whole portfolio's data.

Runs the whole upsell journey around every booking.

Builds the optimal cleaning route from arrivals/departures/stayovers, predicts linen needs, logs maintenance tickets from guest messages, and triages them with context — a VIP arriving in five days bumps that room's repair up the list.

Detects payment disputes/chargebacks, assembles the evidence packet (booking, comms, policy, signatures), and drafts the response within the deadline.

Argues every order from the forecast, not habit.

Handles concierge requests end-to-end: books restaurant tables, private chefs, and airport/VIP transfers on the guest's behalf.

Builds each rota from forecast occupancy, the real room-turn workload, the restaurant book, and your house rules — team composition, legal and contract limits, fairness, cost.

Runs your pricing the way the big-league revenue systems do — continuously, not once a day.

Automates the full procure-to-pay cycle: purchase requisitions with budget checks, approval routing per your delegation-of-authority matrix, PO creation and vendor confirmation, goods-receipt matching, 3-way invoice match with reason-coded mismatch flags, and payment-batch preparation — with a complete audit trail at every step.

One system that makes every VIP feel known.

Scans inboxes and Drive for invoices/receipts, reads each PDF, codes it to the right ledger category, matches it to its purchase order, files it to the right folder with a clean name, and logs it to the finances sheet, automatically.

Tracks energy, water, waste, and laundry volumes per occupied room, benchmarks month over month, drafts the ESG report sections, and flags anomalies (a spike in water use on floor 3) to engineering.

Watches Google, TripAdvisor, Booking.com, and Vrbo reviews and drafts a personalised, on-brand reply to each that thanks the guest, addresses complaints, and pulls in booking context.

Predicts no-shows and late cancellations per night from booking history and lead-time patterns, recommends a safe controlled-overbooking level, and preps the walk plan (which booking, which partner hotel) if the night runs over.

Owns the whole event once someone wants to hold it here — weddings, conferences, offsites, board meetings.

Reads every inbound guest message the moment it lands, classifies intent (booking inquiry, FAQ, arrival time, modification, preference, thank-you…) and manages the three core booking types end-to-end: rooms in the PMS, tables at the restaurant, and treatments at the spa.