Date of Award

6-18-2024

Document Type

Thesis

Publisher

Santa Clara : Santa Clara University, 2024

Department

Computer Science and Engineering

First Advisor

Yi Fang

Abstract

This project introduces a novel Conversational Menu Assistant leveraging Retrieval Augmented Generation (RAG) techniques within a modified Large Language Model (LLM) framework to enhance dining experiences by providing personalized menu assistance. The main innovation lies in the system’s ability to integrate up-to-date menu information from various sources, including web scraping, into its responses, thereby circumventing the limitations commonly associated with LLMs, such as the need for frequent retraining and the challenge of handling dynamic information. Our solution addresses the pressing issue of reducing the workload on restaurant servers and streamlining the ordering process by offering precise menu details, personalized recommendations based on dietary preferences, and an intuitive user interface for an improved customer experience.

The implementation of our Conversational Menu Assistant demonstrates the efficacy of RAG techniques in real-world applications, showcasing a significant advancement over existing LLMs by focusing on restaurant menus. Through a comprehensive development approach utilizing Python for AI and data processing, React for dynamic user interface design, and OpenAI’s API enhanced with menu-specific information, we aim to achieve high accuracy in responses. Preliminary results indicate a positive impact on the dining experience, offering a proof of concept with potential for future expansion to include broader restaurant selection assistance.

Acknowledging potential ethical and societal implications, our project includes mitigation strategies to address concerns such as the impact on server employment and tipping practices. Future work will focus on refining the LLM’s accuracy, particularly regarding dietary restrictions and allergies, and exploring the scalability of our approach to a wider array of restaurants. This project represents a significant step forward in the application of RAG techniques to improve service industry efficiency and customer satisfaction.

Download

Share

COinS