Enhancing Comprehension with LLM, TTS, and RAG: Transcription of Text into Podcasts and Chatbots
Arya Jalindar Kadam, Prasad Rajaram Kute, Chinmay Ashok Kale, Kushal Bhoraji Pathave, Chaitali Shewale, Pawan Wawage · 2025
The rapid growth of Large Lan-guage Models, Text- to-Speech technologies, and Retrieval-Augmented Generation is bringing a new mode of accessing and understanding textual information. This work talks about how these technologies can be coordinated so that texts are converted into conversational podcasts and chatbots. This makes information accessible and interesting to interact with. The system is based on using Llama3.2:3b for generating script, Suno/Bark TTS model for audio synthesis as close to reality, and a RAG-based chatbot for contextual interaction. with users. User studies present considerable changes in understanding and participation, meaning that these machines are going to flip the learning settings. Index Terms—Large Language Models (LLM), Text-to-Speech (TTS), Retrieval-Augmented Generation (RAG), Document-to- Podcast, Chatbot Interaction, Comprehension Enhancement