Applying Spring AI Function Calling for a Task-Oriented Chatbot in Supermarket E-Commerce

Tóm tắt

In the era of digital transformation, integrating artificial intelligence (AI) into e-commerce systems has become crucial to enhance automation and customer experience. This study presents a task-oriented chatbot system for supermarket e-commerce using the Spring AI Function Calling framework, enabling context-aware and goal-driven customer interactions. The backend, developed with Java Spring Boot, provides a scalable and modular foundation for integrating AI functions into enterprise web applications. Spring AI serves as the core, connecting to Google Gemini APIs to handle natural language understanding and execute domain-specific operations such as product lookup, order management, and user data retrieval. A React Native mobile application ensures seamless interaction across Android and iOS, while MySQL manages structured data efficiently. Experimental results show that the proposed chatbot reduces response time, improves accuracy, and automates customer service workflows effectively. This research highlights the potential of combining Spring AI Function Calling with modern backend and mobile technologies to build intelligent, extensible assistants for e-commerce platforms.