Design and Evaluation of a Conversational AI Framework for Real-Time Museum Exploration and Visitor Engagement
Aryan Sah1, Sewank Nande2, Sadiya Shaikh3, Parth Shinde4, Rajveer Yadav5, Siddharth Baunthiyal6 Department of Computer Engineering,
School of Engineering and Technology, D.Y. Patil University Ambi, Pune, Maharashtra, India
Abstract—We introduce a novel conversational AI system that transforms how museum visitors access information and purchase tickets. Our solution leverages GroqCloud’s Llama 3.3- 70B Versatile model to interpret visitor inquiries and generate natural responses, while incorporating Google Custom Search API to retrieve verified information and utilizing Crawl4AI technology for extracting real-time data when needed. The system features an intelligent workflow that analyzes query intent, maintains conversation context, conducts targeted information searches, and facilitates secure UPI-based transactions. Through careful orchestration of these components, we’ve created a platform delivering precise, relevant information with minimal delay. Testing reveals the system significantly outperforms conventional information services, offering accurate responses to complex queries while streamlining the ticket purchasing process. This implementation addresses persistent challenges in visitor services at cultural institutions by reducing wait times and democratizing access to information. Our ongoing development roadmap includes enhanced language support, improved memory persistence for personalized interactions, and robust analytics capabilities to help museums better understand visitor engagement patterns.
Index Terms—AI-powered chatbot, Museum visitor experience, Automated ticket booking, GroqCloud API, Llama 3.3-70B Versatile, Google Custom Search API, Crawl4AI, Real-time web scraping, UPI-based payment processing, Context-aware responses, Multilingual Support
