AI-based copilots are popping up everywhere. But have you ever wondered how to approach building functionality that can understand and transform a user’s intent? In this presentation, we will unveil the process of creating a generative copilot and explore what to do if we want to go beyond prompting with a commercial API. As a practical illustration, we'll examine the development of copilot functionality for visual-based chatbot building solutions. The emphasis will be on how to fine-tune both commercial and open-source LLMs. We'll also delve into assessing the copilot’s performance - highlighting the balance between accuracy and hallucinations which may be an important factor when deciding for a deployment that may break the bank. Emanuel is driving AI innovation at Infobip with a current focus on Generative AI and analyzing the impact of Large Language Models. He joined Infobip after almost a decade working as the Operations Area Manager and Key Researcher of the FAIR AI team at the Know-Center, one of Europe's leading research centers for trustworthy AI, and Graz University of Technology. He received his PhD in Computer Science from Graz University of Technology and his M.Sc. as well as B.Sc. in Software Engineering and Information Systems from the Faculty of Electrical Engineering and Computing at the University of Zagreb. He is a former Marshall Plan fellow and has been working as a visiting researcher at the Computer Science department of the University of California, Los Angeles (UCLA). He is the special issue editor of Frontiers in Big Data - Recommender Systems section, and his main research interests are in the Information Retrieval field of Recommender Systems and a specific focus on algorithmic accuracy, real-time performance, privacy, fairness and biases. His research on fairness in AI and bias in recommender systems was awarded with the Mind-the-gap gender and diversity award of Graz University of Technology in 2022.