EmoTa is an open-access Tamil Speech Emotion Recognition dataset with 936 utterances from 22 native speakers, covering five emotions (anger, happiness, sadness, fear, and neutrality). It supports emotion classification tasks and advances Tamil language processing.

An MCP server for grounded Tamil word-grammar (சொல் இலக்கணம்) analysis — classify native vs. borrowed words, find roots and meanings, and suggest pure-Tamil equivalents. Answers are anchored in authentic sources (Tholkappiyam) and self-enrich from evolving Tamil data. Built for native speakers.