Diego Ordóñez-Camacho, Rafael Melgarejo-Heredia, Mohsen Abbasi, Lucía González-Solis

Proceedings of the 28th World Multi-Conference on Systemics, Cybernetics and Informatics: WMSCI 2024, pp. 81-84 (2024); https://doi.org/10.54808/WMSCI2024.01.81

Diego Ordóñez-Camacho
Facultad Internacional de Innovación PUCE-Icam, Pontificia Universidad Católica del Ecuador, Quito, Ecuador

Rafael Melgarejo-Heredia
Facultad Internacional de Innovación PUCE-Icam, Pontificia Universidad Católica del Ecuador, Quito, Ecuador

Mohsen Abbasi
Facultad Internacional de Innovación PUCE-Icam, Pontificia Universidad Católica del Ecuador, Quito, Ecuador

Lucía González-Solis
Facultad Internacional de Innovación PUCE-Icam, Pontificia Universidad Católica del Ecuador, Quito, Ecuador

Cite this paper as:
Ordóñez-Camacho, D., Melgarejo-Heredia, R., Abbasi, M., González-Solis, L. (2024). Aurel_AI: Automating an Institutional Help Desk Using an LLM Chatbot. In N. Callaos, E. Gaile-Sarkane, N. Lace, B. Sánchez, M. Savoie (Eds.), Proceedings of the 28th World Multi-Conference on Systemics, Cybernetics and Informatics: WMSCI 2024, pp. 81-84. International Institute of Informatics and Cybernetics. https://doi.org/10.54808/WMSCI2024.01.81

The Aurel_AI research project focuses on creating a virtual help desk for universities, delivering accurate information about academic programs, regulations, processes, and personnel to both internal and external clients. Traditional call centers often grapple with outdated data, limited knowledge, and high staff turnover, leading to inaccurate responses and long wait times. Generative AI models, particularly Large Language Models (LLMs), offer a promising solution for automated help desks. These models can comprehend poorly structured queries and generate appropriate answers. However, they may encounter “hallucinations” due to insufficient training data. Ensuring accurate and comprehensive information involves specific data collection, validation, and updating methodologies. Techniques like Fine-Tuning and Retrieval-Augmented Generation (RAG) are essential for specific use cases. While both methods have pros and cons, balancing cost-effective infrastructure is crucial for a precise, flexible, and user-friendly system.