Convin.ai Research Accepted at ICASSP 2026
Chennai: Bengaluru-based AI startup Convin.ai has secured acceptance for its research paper at ICASSP 2026, one of the world’s leading conferences in speech and signal processing, organised by IEEE.
The paper, titled “IPACue-TTS: Integrating Prosody and Articulatory Cues in Conditional Flow Matching for Multilingual Zero-Shot TTS,” will be presented on May 7 in the Speech and Language Processing session.
The research proposes a new approach to multilingual text-to-speech (TTS), enabling natural, human-like voice generation across languages without requiring speaker-specific training data. The work was authored by Giridhar Pamisetty and Atul Shree of Convin.ai.
ICASSP, the flagship conference of IEEE’s Signal Processing Society, is known for its highly competitive selection process and global participation from leading universities and technology firms.
Commenting on the achievement, Atul Shree, Co-founder and Director, said the recognition reflects the team’s efforts in solving real-world challenges in multilingual speech AI with limited resources.
Convin.ai develops AI-driven conversation intelligence tools used by enterprises in sectors such as banking, insurance, and healthcare. Its platform supports voice automation, real-time agent assistance, and post-call analytics, handling millions of customer interactions monthly.
Founded by IIT alumni, the company has built a proprietary AI stack supporting multiple Indian languages and works with clients across India and international markets.
