Worked with Enuncify, a US-based startup, to develop a multilingual, multi-accent Text-to-Speech (TTS) system capable of switching accents within a sentence (e.g., American English and Indian accents). Explored models like Coqui XTTS, AI4Bharat IndicTTS, and Parler-TTS, and built a custom language detection system for better word-level accuracy. Overcame challenges like word repetition and accent inconsistency. Proposed future solutions using Knowledge Distillation and Dynamic Voice Modulation. Grateful to Prof. Aruna Malapati and the Enuncify team for the mentorship!