Agentic AI for Automated Video Summaries and Subtitle Creation

This paper introduces an agentic artificial intelligence framework designed for the automated analysis, summarization, and multilingual subtitle generation of video content. The system integrates advanced large language models (LLMs) from Groq, leveraging their capabilities for high-speed transcription, comprehensive insight extraction—including key topics, actionable items, critical questions, and emotion breakdowns—and robust multilingual translation. Complementing these capabilities, the framework incorporates ElevenLabs for high-quality text-to-speech summary playback and a Streamlit-based interactive user interface. This holistic tool efficiently processes videos from diverse sources (uploaded files or YouTube URLs), generating accurate transcripts, insightful summaries, and dynamically translated subtitles in languages such as English, Hindi, and Kannada. The proposed framework holds significant promise for enhancing content accessibility and streamlining information digestion across educational, media, and business sectors.