Smart Technical Support System Development Using Knowledge Map-Aided Approach

Strong technical support is crucial to success of any brand. It provides answers to customer's questions and solutions to different situations and is an important factor in keeping customers loyal. A new level of technical support can be achieved through combination of traditional service automation offers with the current advancement of Generative Artificial Intelligence (GenAI) embedded into technical support. GenAI in general and the large language models (LLM s) in particular create new opportunities for natural conversational interfaces and the development of ‘smart’ technical support systems. Based on experiences and observations from other intelligent assistance projects, this paper presents new methodological perspectives from academia and best practice from industry on architecting intelligent technical support systems. It discusses the impact of GenAI and LLMs through real cases supporting an ongoing validation. The maj or focus of this paper is on architectural models for intelligent technical support systems, showing the fundamental Knowledge Map mechanism of AIbased customer service, which allows for not only controlling the information used to generate responses and minimizing errors but also automatically processes requests in accounting systems. An example of successful implementation of Knowledge Mapbased approach is given, where the intelligent assistant chat-bot demonstrates high accuracy of answers and in case of uncertainty redirects requests to human operators. The paper presents the key architectural model perspectives for the development of intelligent technical support systems.