Building Indian speech datasets across multiple languages, with a focus on speaker diversity, real-world conversational data, and structured metadata for ASR and voice AI use cases.