Mitesh Khapra: Pioneering AI for Indic Languages and Recognized Innovator on the Global Stage
Celebrating Mitesh Khapra: A Pioneer in AI for Indic Languages
When you think of artificial intelligence (AI), English often springs to mind as the predominant language of innovation. From Alan Turing’s groundbreaking work to the advancements in machine learning today, the Western narrative has largely shaped the AI landscape. However, India, with its rich tapestry of languages and cultures, is on a mission to carve out its own identity in the field—especially for Indic languages.
Enter Mitesh Khapra, an Associate Professor at IIT Madras and a key innovator at AI4Bharat. His groundbreaking work in deep learning, natural language processing, and conversation systems is aimed at revolutionizing AI specifically for the diverse linguistic needs of India.
A Global Recognition
Khapra’s inclusion in the 2025 TIME100 AI List—a compilation of the world’s most influential figures in artificial intelligence—has stirred excitement across the nation. He’s featured alongside titans of the tech world like Elon Musk, Sam Altman, Jensen Huang, and Mark Zuckerberg, but what sets Khapra apart is that he is an Indian innovator actively addressing local challenges.
The profile on Khapra in TIME deeply resonates with the Indian tech community. The magazine highlighted how nearly every Indian startup engaged in voice technology relies on datasets created by him and his team. "The reason Indian language technology is behind English," Khapra explains, "is because we do not have enough data for Indian languages."
Bridging the Gap for Indic Languages
While existing AI models may excel in languages like Hindi and Bengali, they struggle with many of India’s lesser-represented tongues. To address this, Khapra’s research lab, AI4Bharat, conducted an ambitious project that traversed almost 500 of India’s 700 districts. This project aimed to collect thousands of hours of voice data from individuals across varied educational and socioeconomic backgrounds, capturing all 22 of India’s official languages.
Founded in 2019, AI4Bharat has become a key partner in the Bhashini program, an initiative under the auspices of the Digital India Initiative. Bhashini is designed to provide AI-powered digital services in local languages, ensuring that no one is left behind in the AI revolution.
A New Dawn for AI in India
TIME noted that AI4Bharat contributes approximately 80% of the datasets used in the Bhashini program. This open-source initiative not only benefits local startups but also tech giants like Meta and Google, who leverage these datasets to enhance their AI models for languages like Hindi and Marathi. Khapra’s work ensures that technological advancements are inclusive and beneficial for everyone in India.
A Moment of Pride for India
Mitesh Khapra’s journey is a testament to the transformative power of language in the realm of technology. By placing emphasis on data that represents India’s multilingual landscape, he is nurturing an ecosystem where all languages have a place in the world of AI.
As AI continues to evolve, Khapra’s vision for an inclusive technological future is not just aspirational; it is becoming a reality. His work marks a significant step towards empowering millions of Indians to access AI-driven services in their own languages, thus redefining the narrative around AI in India.
In celebrating Khapra’s achievements, we recognize a moment of global pride for India in the expanding universe of artificial intelligence. Through initiatives like AI4Bharat, we are not just bridging technological gaps but weaving a narrative that places Indian languages at the forefront of AI innovation.