In a landmark move that reimagines the future of voice-enabled AI, United We Care, the AI-first mental health pioneer, has announced the launch of Shunya Labs — a deeptech spinout built to redefine the infrastructure of speech recognition.
Billed as a “from-zero” breakthrough, Shunya Labs.AI offers a next-gen Automatic Speech Recognition (ASR) engine that crushes Big Tech benchmarks across speed, accuracy, and multilingual fluency — and does it all on CPUs, making it more cost-effective, scalable, and accessible for enterprise use.
“We didn’t set out to beat benchmarks. We set out to build what didn’t exist,” said Ritu Mehrotra, Founder of United We Care. “We’ve created an AI voice infrastructure that listens like a human, operates like a machine, and prioritizes privacy as a core principle.”
Multilingual, Privacy-First, and Bharat-Ready
Unlike conventional ASR tools, Shunya Labs is tailored for India’s polyglot ecosystem. It natively understands over 32 Indic languages, including Hindi, Bengali, Kannada, Assamese, and Maithili — with seven more in the pipeline. The engine powers everything from rural health kiosks to call centers across India and beyond.
Key performance highlights include:
3.37% Word Error Rate — setting a new benchmark for speech recognition accuracy
20× reduction in cloud infrastructure costs — with optimized CPU-based deployments
On-prem and edge-ready deployment — enabling real-time transcription in low-connectivity settings
Military-grade data privacy and explainability — ensuring enterprise-grade compliance and trust
Originally designed to support Stella, United We Care’s AI wellness engine, Shunya Labs has already shattered nine global records in speech and language understanding. Innovations like a 230M+ node Clinical Knowledge Graph and Spatio-Temporal Graph Attention Networks (STGAT) form the foundation of its emotionally intelligent and linguistically nuanced capabilities.
“Shunya isn’t just a name. It’s our origin story,” said Sourav Banerjee, Co-founder & CTO. “Just as zero changed mathematics, we’re building a new base layer for AI speech — engineered from first principles, not APIs.”
Built for the World’s Most Demanding Environments
From hospitals and banks to defense systems and public sector deployments, Shunya Labs delivers real-time, secure voice AI that thrives in mission-critical scenarios.
As it gears up for an upcoming release on Hugging Face and production rollouts across continents, one message stands out: Voice is not just a feature. It’s infrastructure.
And with Shunya Labs, that infrastructure is proudly built in India — and built to last.
