Overview:
Global Travel Collection (GTC), a luxury and corporate travel advisory network, developed Atlas, a generative AI-powered platform to streamline complex trip planning. Traditionally, advisors spent six to eight hours per itinerary manually searching across 10+ disparate travel systems, leading to inefficiencies and lost productivity. Atlas centralizes travel content, delivering contextual answers via natural language, reducing manual lookup and accelerating advisor workflows. The solution operates across travel planning processes and integrates enterprise data sources to provide actionable insights for itinerary creation.
The platform is built on Microsoft Azure AI Foundry and Copilot Studio, leveraging Azure AI Search, Azure OpenAI Service, and other Azure services to drive retrieval-augmented responses and orchestrate AI workflows securely and at scale. Atlas is designed to meet modern user expectations for conversational search while maintaining governance, compliance, and security required for handling sensitive traveler data.
Key Features:
- Unifies travel data from multiple airline, hotel, and entertainment sources into a single AI-enabled interface.
- Allows advisors to input conversational natural language queries and receive context-rich responses.
- Deploys retrieval-augmented generation (RAG) using Azure AI Search and Azure OpenAI to fetch and generate grounded travel information.
- Provides enterprise governance, centralized model orchestration, and secure index management via Azure AI Foundry.
- Offers a low-code interface with Copilot Studio for rapid development, testing, and refinement of AI features.
- Supports session context and advisor interaction tracking through Azure Cosmos DB for future personalization and analytics.
Results & Impact:
- Reduces travel planning time by about three hours per itinerary on average.
- Projected to reclaim more than 1.5 million advisor hours annually across GTC’s ~1,700 advisors.
- Transforms advisor queries that once took 10–15 minutes into seconds with AI-powered search.
- Improves advisor experience by consolidating fragmented travel resources into one intuitive system.
- Establishes a platform for future monetization opportunities, such as partner integrations and premium services.
AI Technology:
AI Model Types: Generative AI, large language models (via Azure OpenAI Service), retrieval-augmented generation.
AI Purpose: Summarize and synthesize multi-source travel data, conversational query understanding, and automated information retrieval.
Application Type: Operations and knowledge management for travel planning and advisory support.
Target Users:
- Travel advisors specializing in luxury and corporate itineraries.
- Operations and product management teams at GTC and Internova.
- IT and AI governance teams responsible for secure model deployment.
- Partner travel and vendor relationship managers exploring integration opportunities.
Sources:
- microsoft.com/en/customers/story/23943-global-travel-collection-azure-ai-foundry