🔥36氪•較早收集於 25m
飛豬推出酒店商家專用 AI 智慧助理
💡這是利用 LLM 簡化非技術商業用戶複雜後端操作的絕佳案例。
⚡ 30-Second TL;DR
有什麼變化
透過自然語言指令管理庫存與價格
為什麼重要
此工具顯著降低了酒店管理者的營運負擔，並透過 AI 洞察提升了服務品質。
下一步行動
探索如何將類似基於 LLM 的「意圖轉行動」介面整合到您自己的 B2B SaaS 產品中。
誰應關注：Enterprise & Security Teams
關鍵要點
- •透過自然語言指令管理庫存與價格
- •提供定價策略的即時數據分析
- •自動優化旅客入住全流程的服務品質
🧠 深度解析
Web-grounded analysis with 13 cited sources.
🔑 增強重點摘要
- •Fliggy's AI strategy employs a multi-agent system, where specialized AI agents collaborate to handle complex tasks, emulating the workflow of human travel consultants for both consumer-facing and B2B scenarios.
- •The AI assistant is deeply integrated with Alibaba's Qwen AI models and utilizes Fliggy's proprietary, high-quality travel scenario datasets, encompassing live pricing, inventory status, travel routes, and attractions.
- •Beyond direct merchant tools, Fliggy has introduced 'flyai,' a developer-facing travel skill available on platforms like ClawHub and GitHub, enabling external developers to integrate Fliggy's extensive travel inventory and services into their own AI-powered applications.
- •Fliggy's AI interface has evolved from merely planning to full booking fulfillment, processing reservations directly within Alibaba's Qwen app, which led to an 800% increase in overall AI orders and a 24-fold jump in attraction ticket bookings during the Spring Festival.
- •The scope of Fliggy's AI applications extends beyond hotel management to include supply chain operations, customer service, platform governance, product development, and destination marketing, with approximately 10% of online customer inquiries managed by AI as of late March 2025.
📊 競品分析▸ Show
| Feature / Company | Fliggy AI Assistant (Hotel Merchants) | IDeaS | Duetto | SiteMinder | Jurny |
|---|---|---|---|---|---|
| Core Function | Hotel inventory, pricing, invoice management, service optimization via natural language. | AI-driven revenue management, dynamic pricing, demand forecasting. | AI-powered revenue management, dynamic pricing, demand forecasting. | Hotel commerce platform, global distribution, real-time rate management, dynamic pricing. | AI-powered platform for short-term rental & hotel management (guest comms, reservations, revenue, housekeeping). |
| AI Approach | Multi-agent system, Alibaba Qwen AI models, proprietary datasets. | Machine learning, deep analytics for pricing optimization. | AI-powered algorithms for dynamic pricing and demand forecasting. | AI-driven tools for channel identification and dynamic pricing across channels. | AI-driven insights, automated review responses, unified inbox. |
| Integration | Deep integration with Alibaba ecosystem (Qwen app, Taobao, Alipay, Amap). | Integrates with PMS, CRS, channel managers. | Integrates with PMS, CRS, channel managers. | Connects with global distribution channels, PMS, CRS. | Unified data layer connecting PMS, CRM, third-party apps. |
| Target Market | Hotel merchants, B2B travel, developers (via flyai). | Hotels of all sizes, from independent to global chains. | Large hotel groups and resort chains. | 41,000+ hotels across 150 countries. | Short-term rentals and hotels. |
| Pricing Model | Not publicly disclosed. | Not publicly disclosed. | Not publicly disclosed. | Not publicly disclosed. | Not publicly disclosed. |
| Benchmarks/Performance | 800% increase in AI orders, 24-fold increase in attraction ticket bookings during Spring Festival. | Companies leveraging AI in revenue decisions saw 5-15% revenue improvement. | Reduces support ticket volume by 40% or more (for guest experience platforms). | N/A | N/A |
🛠️ 技術深入
- AI Model Integration: The AI assistant is integrated with Alibaba's Qwen AI models, which are large language models (LLMs).
- Multi-Agent System: It employs a multi-agent collaboration system, where specialized AI agents autonomously break down tasks, plan workflows, and deploy sub-agents for specific functions like flight search, hotel selection, or budgeting. This design mimics human travel consultant workflows.
- Data Training: The system is trained using Fliggy's proprietary, high-quality travel scenario datasets, which include real-time information on flights, hotels, attractions, inventory status, and pricing.
- Real-time Data Engine: It is integrated with Fliggy's real-time pricing engine to provide accurate and up-to-date travel information and bookable options.
- Ecosystem Integration: The AI interface deeply integrates with core services across the broader Alibaba ecosystem, including the Qwen app, Taobao, Taobao Instant Commerce, Alipay, and Amap, to orchestrate search, product selection, payment, and related services.
- Developer Tool (flyai): The 'flyai' skill is built on Fliggy's Model Context Protocol (MCP) standard, designed to be plug-and-play for developers to integrate travel inventory and services into AI agent workflows without building booking infrastructure from scratch. It is available on ClawHub and GitHub.
🔮 前景展望AI analysis grounded in cited sources
AI-native booking experiences will become a dominant consumer expectation in travel.
Fliggy's reported 800% increase in AI orders and 24-fold jump in attraction ticket bookings during the Spring Festival, coupled with integration into smartphone ecosystems, indicates a strong consumer shift towards AI-orchestrated travel transactions.
Fliggy will expand its role as a foundational travel infrastructure provider for other AI applications.
The launch of 'flyai' as a developer-facing tool, enabling third-party integration of Fliggy's travel inventory into various AI agent workflows, positions Fliggy to power a wider ecosystem of AI-driven travel services.
The multi-agent AI architecture will set a new industry standard for delivering complex, personalized services in hospitality and travel.
Fliggy's emphasis on specialized AI agents collaborating to emulate human expertise for personalized planning and operational tasks demonstrates a scalable and effective model for sophisticated service delivery.
⏳ 時間線
2016-10
Alibaba launched Alitrip, later rebranded as Fliggy.
2017-08
Alibaba Group and Marriott International announced a strategic co-operation, with Fliggy hosting a Marriott hotel flagship store.
2025-04
Fliggy launched 'AskMe,' its smart AI travel assistant, powered by a multi-agent system and Alibaba's Qwen AI models.
2025-06
Alibaba merged Fliggy into its core e-commerce business to streamline operations and sharpen its focus on AI.
2025-09
AskMe introduced new features including itinerary maps, heat-based maps, and photo-based audio AI guides.
2026-03
Fliggy went live with end-to-end bookings via its AI interface, processing reservations through Alibaba's Qwen App, and launched 'flyai' for developers.
📎 來源 (13)
Factual claims are grounded in the sources below. Forward-looking analysis is AI-generated interpretation.
📰
AI 週報
閱讀本週精選 AI 大事摘要 →
👉相關動態
AI 策展新聞聚合。所有內容版權歸原始發布者所有。
原始來源： 36氪 ↗