AI-based Analysis of Service Notifications Prototype

This document is a technical report detailing the development of a prototype for analyzing service notifications using AI technology. The initial situation highlights the challenge of managing hundreds of thousands of service notifications from customers each month, which include unstructured data in multiple languages. The report outlines the necessity of understanding common issues and solutions identified by technicians in the field, which is not feasible through manual analysis due to resource constraints. The proposed solution involves leveraging OpenAI's large language model (LLM), specifically GPT-4, to automate the analysis process. The prototype is designed to read service notifications, generate structured data, and cluster similar topics without human intervention. This innovation allows for the first-time use of data in further product quality analyses, enhancing the efficiency and effectiveness of the service notification process.