India’s doctor patient ratio stands at 1:1000 Indians, way below the standards prescribed by the World Health Organisation. Clinical documentation plays a critical role in facilitating better patient care with seamless billing, insurance, medico-legal processes. COVID has forced healthcare providers to deploy highly efficient medical reporting systems to minimise human contact and allow doctors to have dynamic working hours. The pandemic has pushed the widespread adoption of electronic medical records.
Rustom Lawyer founded a cloud-based speech-to-text software, Augnito, in 2020 that guarantees error-free medical documentation by converting human voice to written text in real-time. Augnito is India’s first and only medical speech recognition AI available as a real-time speech-to-text application for doctors to generate reports. Rustom is also the man behind Scribetech, a 19-year-old company that pioneered clinical documentation in the UK NHS.
Analytics India Magazine got in touch with Rustom Lawyer, the co-founder & CEO of Augnito, to gain insights into the company’s processes.
How Does It Work?
A truly ‘made in India’ product, Augnito uses Human-Centred Design principles to understand the needs and challenges of doctors unique to the Indian context. It took seven years for the company to develop this software. “Currently, the potential use cases are very exciting, and we are the only company in India providing such an advanced voice technology in the healthcare domain,” said Rustom.
The software has been built on deep neural network-based technology that leverages the latest and most cutting edge techniques in speech recognition science. The strong natural language processing team and robust NER (Named entity recognition) techniques help structure the text. “We also have a large in-house GPU infrastructure,” said Rustom.
With a cloud-based solution built on highly advanced deep learning models, Augnito offers the highest accuracy rates of 99% for diverse Indian accents without any voice training. “It’s like you can carry the entire language of medicine with you, everywhere. The flexibility and mobility that Augnito offers by being a cloud-based multi-platform solution is one of the key differentiators” added Rustom.
To facilitate this on the cloud, the developers use tools like Kubernetes, Python and Amazon DynamoDB and on the front end, and Angular, Javascript and Dot Net tools for the backend. The system is available as a streaming speech API for direct integration and contains the entire language of medicine covering 50+ specialities and subspecialties.
The system has also been recently awarded the prestigious Abdul Kalam Fellowship by the Government of India.
Wrapping Up
Rustom believes speech recognition and natural language processing are currently at an exciting phase and can be a critical factor driving electronic medical records adoption. With Augnito, Rustom tried to package this technology into a first-of-its-kind system for our country, easily accessible to every doctor.
Augnito has invested about a million dollars in the business (completely bootstrapped) and is currently being used by 4000+ doctors across India to produce medical reports. “In the next five years, the startup is expecting to see Augnito as the voice interface driving software in every hospital across the country,” concluded Rustom.
