AI-ENHANCED
HOSPITALITY
DATA
PLATFORM

Our client is a globally recognized platform aggregator of data from hotels worldwide to support corporate event management. They offer insights into market analytics, industry trends, and benchmarking tools for the hospitality industry. Their technology empowers hoteliers to define market potential, target profitable opportunities, increase market share, develop new customers, and understand booking patterns.

Their suite of hospitality solutions harnesses the power of data and analytics to help customers know their buyers and succeed in the group and meetings marketplace. With the industry’s largest historical database of actualized events, they deliver insights that increase market share, optimize revenue strategies, and accelerate sales cycles.

• Stay on the technological edge by leveraging external consulting and expertise to implement AI solutions, ensuring competitive advantage and enhanced data operations.
• Identify specific business areas, that can be significantly improved with AI solutions to drive better user experiences and operational efficiency.
• Develop a detailed roadmap for key software solutions, integrating AI to enhance their capabilities.
• Focus on improving data processing and handling diverse information sources, ultimately leading to more accurate insights and decision-making.

Svitla provided comprehensive consulting services to analyze the client’s business and suggest possible digital transformation strategies through AI solutions.

Consequently, four proof-of-concept (POC) approaches were developed to strengthen existing solutions and provide a competitive edge.

• AI-Based Search Functionality: Svitla Team upgraded the existing search solution, which relied on a limited number of preset filters, to an AI-driven search system that interprets individual user language and generates reports based on customized requests. The new system provide more detailed and specific results. This enhancement allows for tailored results, complex queries, and effective searches across a broad spectrum of data. The system was built using GPT as the main model, integrated with Elastic Search for precise data retrieval, AWS Lambda and API Gateway for backend services, and Python for the core code.
• Demand Forecasting: Svitla Team addressed the client’s need for accurate forecasting service, to enable users plan and manage event quantities effectively. The solution was based on Prophet by Meta for its robust forecasting capabilities, utilizing Streamlit and Python for the backend code.
• Recommendation Engine: The Svitla Team developed a recommendation engine to help users find relevant companies with similar characteristics, enhancing the overall user decision-making process. The system was built using Python for the core code, KNN for the main algorithm, pandas for data processing, and Streamlit for the web app.
• AI-Driven Data Extraction from Print Banners: Svitla Team developed a GPT-4 Vision-based service to streamline the conversion of printed event banners into digital format. The solution utilized Python for core development, Streamlit for the web app, GPT-4 Vision for data extraction, and pandas for processing and structuring the extracted data.