We have developed a Retrieval Augmented Generation Large Language Model specifically designed for IT help desks. The proposed model aims to enhance the efficiency and accuracy of the responses provided by engineers when addressing technical queries related to various IT topics, including but not limited to Red Hat, Windows, Azure, Kubernetes, and Cisco. A dedicated knowledge base of IT-related articles, documentation, and troubleshooting guides is meticulously maintained. This knowledge base functions as a search component, enabling the model to present relevant information based on user queries. For an IT help desk application to gain user trust, it must be supported by appropriate techniques and tools. Furthermore, when provided resources, such as manuals or books, lack the necessary information, the language model must transparently respond with “I do not know”. This honesty fosters a climate of trust and ensures that users receive accurate and reliable assistance. Performance evaluation metrics for the model include precision and recall, facilitating the assessment of its effectiveness in real-world applications, such as IT support. Encouraging results have been achieved, and further improvements are anticipated by expanding the knowledge base to include additional IT domains. This approach not only aims to refine the model’s capabilities but also strives to provide comprehensive support across a broader range of technical issues, thereby enhancing overall user satisfaction and operational efficiency in IT service environments.