Voice-based analytics help enterprises to search for data and generate reports. This solution is based on speech recognition. The reports and results can be presented in different formats as per commands. The voice commands are based on natural language and keyword-based queries. The text-based queries are typically different from voice query length. Voice queries may not be questions but more as commands to search and generate results. The speech recognition and processing happen in steps like coding, synthesis, and recognition. The speech-based queries and commands can be connected or isolated words. The speech can be continuous and spontaneous. The speaker models help in identifying person-specific speech patterns. Speaker specific speech models are accurate and precise for voice-based analytics. Diarisation helps in partitioning the speech into homogeneous segments specific to speaker.