<![CDATA[The rapid advancement of digital technology has increased the demand for information delivery systems that are accurate and contextually relevant. This study aims to design and evaluate the performance of a LLaMA3-based chatbot integrated with the Telegram platform using the Flask framework. The development is driven by the need for digital information services capable of providing fast, relevant, and context-aware responses in the Indonesian language. The research methodology includes system configuration, API integration, and performance testing through measurements of Average Response Time (ART) and response quality assessment based on relevance, clarity, and completeness. Experimental results show that the chatbot achieves an average response time of 49.15 seconds and an average response quality score of 14.10 out of 15, indicating strong contextual understanding and high response relevance. These findings suggest that the chatbot system is feasible for use as an automated Indonesian-language information service. However, further optimization is necessary to improve processing speed for broader and more dynamic implementation scenarios.]]>
Copyrights © 2025
