From Sarvam AI to Open Healthcare Network, AI-based solutions to real-world problems

At the Microsoft AI Tour event, the company hosted a select developer showcase to give an overview of how Indian developers are discovering opportunities in AI to solve global challenges; Image: Shutterstock

During the Microsoft AI Tour event in Bengaluru, a select developer showcase was hosted by the company to give an overview of how developers in India are discovering opportunities in AI (artificial intelligence) to solve real-world challenges.

“India is a country with the transformative power of what we can do for some of the most complex challenges when it comes to education, to healthcare, to things which are happening with language barriers, they can really be simplified. And you can go to the grassroot levels to create citizen-scale solutions,” said Irina Ghose, managing director, Microsoft India.

On the skilling front, Microsoft has initiatives such as Code Without Barriers and AI Odyssey, that Microsoft chairman and CEO Satya Nadella spoke about during his keynote speech on Thursday morning. In January, Microsoft enabled 100,000 developers to advance their careers in AI through its AI Odyssey initiative. It allowed participants to take the first step towards becoming a subject matter expert in AI by learning new skills and earning Microsoft credentials.

Related stories

• 

  India's developer community to play a key role in global AI innovation: Satya Nadella

• 

  Microsoft will provide AI skilling opportunities to 2 million Indians by 2025: Satya Nadella

• 

  What the OpenAI saga shows us about Microsoft—and Satya Nadella's—dominance

“On the productivity side, we've really seen that GenAI is taking productivity to the next level. For instance, we are seeing that with GitHub CoPilots, productivity goes up by around 55 percent,” said Ghose while addressing the media. India is the fastest-growing market on GitHub, a Microsoft-owned software collaboration and innovation platform, with 13.2 million developers using the platform. It is expected to overtake the US as the largest developer community on GitHub by 2027.

Here are some of real-world solutions developed by startups built using Microsoft’s solutions:

SARVAM AI

Generative AI startup Sarvam AI—founded by Vivek Raghavan and Pratyush Kumar— has announced that it is working with Microsoft to make its Indic voice large language model (LLM) available on Azure. The startup is known for innovating on all layers of generative AI: Data, models, and apps.

Sarvam AI’s Indic voice LLM, which intends to provide a natural voice-based interface to LLMs, will initially be available in Hindi. Sarvam AI is working to expand coverage to more Indian languages while ensuring support for colloquial language use.

“We are very excited to collaborate with Microsoft to make advanced AI technology accessible and impactful for everyone in India. This partnership embodies our vision of ‘Sarvam’—meaning ‘all’—by enhancing AI’s reach across various Indian languages and sectors,” said Pratyush Kumar, co-founder of Sarvam AI.

Voice is one of the most natural interfaces for generative AI applications in Indian languages and can be applied in sectors such as education, finance, health care, and customer service. The companies will also collaborate to help enterprises adopt generative AI quickly and responsibly. Vivek Raghavan, co-founder of Sarvam AI, said, “We see great synergy in combining the deeptech expertise at Sarvam AI with Microsoft’s leadership in frontier models and the Copilot stack to empower enterprises to be successful in Generative AI.”

During the showcase, the startup demonstrated 'Jan Ki Baat' which is a collaboration with People+AI that shows how telephony-based outbound calls can be made to many people to perform qualitative surveys at scale. The second application was 'Call a Teacher', an application done in partnership with EkStep Foundation where a student studying a chapter in an NCERT textbook can send the QR code of that chapter to a WhatsApp number and get a call from a teacher bot (which is created fully automatically) which will teach this lesson to the student conversationally. This use-case will powerfully establish how frontier models such as GPT-4 can be combined with efficient smaller models such as Sarvam AI's voice LLM to create entirely new experiences to make 'intelligence' more accessible to all.

OPEN HEALTHCARE NETWORK

The Open Healthcare Network (OHC) is a community lead effort that was created to deliver health care in resource constrained settings. In four years, it has become one of India’s most active open source communities, with over 400 contributors from around the world.

But OHC’s transformative impact started with the 10BedICU programme, which was launched at the peak of the Covid pandemic to create ICUs in government hospitals in smaller towns and villages. The CARE platform was the digital backbone of the programme, and “OHC has also built a TeleICU module to connect specialists from medical colleges to help patients in remote 10BedICU wards over a cloud network”, says Srikanth Nadhamuni, chairman, 10BedICU. Today 10BedICUs have been deployed in over 200 government hospitals in nine states covering 23 percent of India’s district, delivering life-saving care to over 15,000 patients.

Also read: Demystifying AI in healthcare in India

Nadhamuni adds, “Microsoft GitHub has played a game-changing role in accelerating OHC's development pace and maintaining the community. Over 400 developers across the world, collaborating remotely, have contributed to OHC's projects under the MIT licence, fostering a new era in open-source, community-driven health care solutions.”

GitHub Copilot’s code generation capability has dramatically enhanced OHC's project development. This has not only propelled efficiency gains but also enabled OHC to swiftly adapt and improve its software offerings to meet the changing demands of the health care sector, as new Gen AI capabilities emerge.

PERSISTENT SYSTEMS

Persistent is the leading technology partner to industry associations focussed on aggregating Indian datasets that will help identify therapies unique to India. The company has set its sights on helping find targeted drugs and treatment protocols for chronic diseases affecting the Indian population.

Persistent enabled Renalytix to commercialise industry-leading chronic kidney disease risk assessment products globally. “The current landscape of disease treatment and drug development is predominantly tailored to Western populations, which means there is a disparity in the effectiveness of treatments for patients in India,” says Pandurang Kamat, chief technology officer, Persistent Systems. “India faces a daunting challenge with a rapidly escalating disease burden. Some of the bigger cohorts are approximately 27 million cancer patients and 70 million plus patients for diabetes.”

One key aspect of creating targeted and effective treatment protocols for the Indian population is to collect and analyse genomic information. The volume of raw genomic data amassed from each patient is immense, surpassing one terabyte, and cumulatively, for large population research, this data will run into Petabytes.

As a solution, Persistent has created a multi-omics platform on Microsoft’s Azure to democratise access to this genomic data through the creation of a data commons; enable highly scalable and secure analytics pipelines and visualisation; identify gene alterations to predict target genes that will aid drug discovery, and identify and validate new biomarkers and pathways of disease progression.