A major step forward for healthcare technology, Shunyalabs.ai has unveiled ZeroMed, a next-generation speech recognition system designed to transform how clinicians document patient care. Built from the ground up for medical use, ZeroMed promises faster, more accurate transcription all while keeping sensitive data fully secure.
Shunyalabs.ai, known for its pioneering work in VoiceAI infrastructure, says ZeroMed delivers “best-in-class accuracy” with a word error rate of just 11.1%, outperforming competitors such as OpenAI’s Whisper V3 and AWS Transcribe. But it’s not just about accuracy, it’s about accessibility. Using a proprietary training process, the company has managed to get the system up and running with only three days of training on two A100 GPUs, a fraction of the time most AI systems need.
“At Shunyalabs, we believe medical transcription must be not just fast, but flawlessly accurate; every dosage, diagnosis and timestamp matters. ZeroMed embodies that vision,” said Ritu Mehrotra, CEO and Founder of Shunyalabs.ai. “We’ve reduced the cost and time to train, making high-fidelity ASR accessible to more healthcare systems.”
Tackling the Real-World Challenges of Healthcare
Medical transcription is notoriously difficult. Conversations in clinics and hospitals are full of overlapping dialogue, jargon, and complex terminology. Even a small transcription error can change a diagnosis or dosage. On top of that, many hospitals face strict privacy regulations that make it hard to use cloud-based AI systems.
ZeroMed tackles these challenges head-on.
It comes with built-in support for medical terminology, drug names, ICD and LOINC codes, and even dosage normalisation, meaning it can automatically format and expand abbreviations that would otherwise need manual editing. The system can also distinguish between different speakers, tracking context in real time even when conversations overlap.
For healthcare providers, that means faster documentation, fewer mistakes and more time focused on patients rather than paperwork.
More from Tech
- National Engineering Day: How The UK’s Engineers Are Making Sci-Fi Tech A Reality
- Aembit Introduces Identity And Access Management For Agentic AI
- Building Resilient IT Support in the Age of Remote Work and AI
- With LTX-2, Lightricks Unleashes AI Video Creation In Full HD
- ChatGPT Wants To Know What Turns You On – Is Desire The Next Dataset?
- Who Owns Your DNA Data? Inside The Legal Grey Zone Of Genetic Data Rights
- Experts Share: How Can The Autumn Budget Better Support Tech Growth?
- From Royals to Researchers: Why Screen-Free Tech Is Becoming More Mainstream
Privacy By Design
One of ZeroMed’s standout features is its on-premises deployment option, which allows hospitals to run the system entirely on local CPU infrastructure without relying on the cloud. That means full compliance with HIPAA, GDPR and other privacy frameworks, a crucial capability for any organisation handling sensitive patient information.
“Our goal with ZeroMed wasn’t incremental improvement, it was to redefine medical speech recognition: fewer corrections, lower latency, and complete data privacy,” said Sourav Banerjee, CTO of Shunyalabs.ai.
Real-Time Innovation
ZeroMed’s “realtime-first” design ensures the same high accuracy whether used live during consultations or for offline transcription. From telemedicine sessions to ambient clinical scribing and radiology reporting, the system has been built to fit seamlessly into modern healthcare workflows.
According to benchmarking data, ZeroMed doesn’t just lead in accuracy, it’s also faster and more flexible than its rivals. Its short training cycle means Shunyalabs can quickly retrain and release updated versions, ensuring the system keeps pace with evolving medical language and new procedures.
What’s Next?
ZeroMed is currently available for preview and pilot evaluation, with early access open to hospitals, telehealth providers, and healthcare startups. It’s initially launching in English, with support for Indian and other international languages expected soon.
As the healthcare sector continues its digital transformation, tools like ZeroMed could mark a turning point, turning medical transcription from a time-consuming necessity into a smart, reliable partner for clinicians.
