Imagine effortlessly gauging how captivated students truly are with online courses. That’s the promise of LLM-SEM, a groundbreaking new metric leveraging the power of large language models (LLMs) like those behind ChatGPT. Traditional methods like surveys often fall short, struggling with small sample sizes and the complexities of interpreting textual feedback. LLM-SEM tackles these limitations head-on. By analyzing video metadata like views and likes alongside the sentiment expressed in student comments, this innovative approach paints a far more accurate picture of engagement. The magic lies in the LLM's ability to decipher the nuanced emotions within those comments, transforming fuzzy feelings into quantifiable data. This data, combined with traditional metrics, allows for a holistic evaluation of engagement at both the course and individual lesson levels. Researchers tested various LLMs, including fine-tuned versions of RoBERTa, and found that LLM-SEM provides a scalable and precise engagement measure. This research opens exciting new possibilities for educators and content creators. Imagine tailoring lessons in real-time based on student sentiment, ensuring maximum impact and a truly personalized learning experience. While the research focused on Arabic-language educational content, future work aims to expand to other languages, bringing the benefits of AI-powered engagement analysis to students worldwide. The challenge now is to refine these models, ensuring they accurately capture the subtle signals of genuine learning. But the journey has begun, and the potential of AI to revolutionize education is clearer than ever.