The creator of 'My Pocket Imam' recognized the challenge of accessing specific textual information from Islamic YouTube videos due to ineffective search capabilities.
To address this issue, a Retrieval-Augmented Generation (RAG) AI system was developed to convert scholarly YouTube halaqas into a searchable knowledge base, providing instant and contextualized access to Islamic knowledge.
The process involved transcribing halaqas, structuring the text into manageable chunks, embedding them for semantic understanding, storing them in a vector database for efficient retrieval, and connecting the system to a chatbot interface for user interaction.
The 'Pocket Imam' chatbot has had a significant impact, offering personalized and immediate insights drawn from scholars' teachings, enhancing the user's spiritual journey by making profound knowledge more accessible and actionable.