While mainstream knowledge is extensively digitized, valuable indigenous worldviews, often embedded in oral narratives, are rapidly eroding due to inadequate representation. To preserve indigenous worldviews, this work leverages Artificial Intelligence to analyze oral narratives. The proposed solution uses a multi-modal approach, sourcing knowledge from audio recordings. An AI pipeline then converts these recordings, often in colloquial language, into a local language text corpus. We made this new corpus accessible for natural language queries using Retrieval Augmented Generation (RAG) and also subjected it to semantic analysis. To highlight the unique indigenous perspective, we contrast the RAG and semantic analysis outputs with those from a baseline corpus of mainstream news stories. The resulting differences in the generated text are then analyzed to reveal the underlying distinctions between the two worldviews. This analytical process serves a dual purpose: it creates a framework for the preservation of this vital cultural heritage. It underscores its importance for the socio-economic development of indigenous communities and the enrichment of global knowledge.