| AI has made a lot of progress lately and a significant factor was the development of Generative AI (GenAI) which has opened the way to a next level of human-machine interaction. An opportunity has been created for companies using GenAI to assist employees in their daily tasks increasing their productivity. This paper introduces a Retrieval- Augmented Generation (RAG) assistant agent designed to handle user queries in natural language regarding business-related documents within the Hellenic Telecommunications Organization S.A. (OTE). The proposed GenAI-driven model provides answer to user queries by using the synthesis of knowledge retrieval and generative capabilities ensuring contextual accuracy. Through the combination of the RAG technology incorporating Large Language Models (LLM) the assistant agent can now generate answers in natural language enhancing both the accuracy of responses and the user experience. By adding Azure Cognitive Search and Azure OpenAI services to the system there is the possibility to integrate domain-specific business rules tailored to support user needs effectively and accurately. The document also displays the challenges that have been identified and addressed during the development of such a system and the models that have been introduced and implemented. It also provides an overview of the system’s performance in a real business environment representing an innovative method of knowledge retrieval efficiencies. |
*** Title, author list and abstract as submitted during Camera-Ready version delivery. Small changes that may have occurred during processing by Springer may not appear in this window.