Utilizing Large Language Models (LLM) for Optimizing Faculty Schedule Communication through Google Calendar Integration

This study presents the development of a Large Language Model (LLM)-based system, specifically LLaMA 3.2 11B, to optimize schedule communication in higher education institutions through integration with the Google Calendar API. The system is designed to understand natural language inputs, detect schedule conflicts, and provide adaptive solutions in real-time. Leveraging advanced LLM capabilities, the system effectively handles ambiguous and multilingual inputs while ensuring contextual accuracy. A dataset of 1,000 entries simulating academic schedules was utilized, including 50% ambiguous queries and 20% multilingual cases, reflecting diverse real-world scenarios. Comparative benchmarks with rule-based and classical machine learning models demonstrate the superiority of the LLM-based approach, achieving a conflict detection accuracy of 96%, latency of 6 seconds, and user satisfaction score of 4.6 out of 5. These results outperform rule-based (84% accuracy, 11 seconds latency) and classical ML systems (90% accuracy, 9 seconds latency). The findings highlight the system’s ability to dynamically adapt to user preferences and real-time scheduling challenges, offering significant efficiency improvements and user-centric interactions. This research contributes to the advancement of AI in digital scheduling systems, emphasizing scalability and applicability in multilingual, high-load environments. Future work includes expanding datasets, enhancing NLP for ambiguous inputs, and integrating cross-time-zone scheduling capabilities.