Earlier this month, tech executives from Google DeepMind, Microsoft, and Meta Platforms converged in Bangalore to witness a major development from one of India’s leading AI startups. Sarvam AI, often compared to India’s version of OpenAI, unveiled a groundbreaking product that could revolutionize the use of AI in the country. Their new software enables businesses to interact with customers through spoken voice in addition to text. Developed using data from 10 native Indian languages and priced at just a rupee per minute, this technology aims to capture a substantial market share. Venture capitalist Vinod Khosla, an investor in Sarvam, highlighted its potential, stating, “These voice bots have the potential to reach a billion people.”
India has been striving to keep up with the global AI boom since the launch of ChatGPT nearly two years ago. However, traditional chatbots have struggled with the limited availability of data for many Indian languages. While urban users might interact with chatbots in English, a significant portion of India lacks this linguistic capability. Startups are now betting on voice bots that use local language data to broaden their reach within India and potentially attract users from other countries.
This burgeoning field is turning India into a testing ground for the next wave of generative AI products, despite some safety concerns seen in other markets. By integrating AI voice capabilities, companies aim to deliver more interactive and conversational services that can respond verbally in real-time and automate various tasks.
Several Indian startups are leading the way. Samsung-backed Gnani AI handles millions of voice interactions daily for major banks, insurance companies, and car manufacturers in India. CoRover AI provides voice bots in 14 Indian languages for the state railway corporation and regional police forces. Haloocom Technologies offers a voice bot that operates in five Indian languages for customer service and job screening.
Ankush Sabharwal, co-founder and CEO of CoRover, noted the significance of voice technology, saying, “The world went from digital first to mobile first to AI first, but voice is the most intuitive way to use technology.” CoRover’s Ask Disha voice bot, launched this month, enables users to book train tickets and make payments entirely through voice commands. Similarly, Gnani AI’s bots assist lenders with financial assessments and help Tata Motors gather feedback and sell automotive accessories.
Sarvam’s voice bots excel in handling mixed-language conversations and performing tasks such as setting appointments and processing payments. The company serves around 50 clients, including the Sri Mandir devotional app, which uses Sarvam’s technology to guide users through rituals and blessings at various temples. Co-founder Vivek Raghavan asserted the advantages of their approach over models like GPT-4, citing the lack of sufficient spoken Indian language data available to US companies.
Despite some hesitation from major AI firms like OpenAI, which has raised concerns about emotional reliance and potential misuse of its voice technology, Indian startups are enthusiastic about their advancements. Ganesh Gopalan, co-founder and CEO of Gnani, emphasized that AI tailored for specific use cases, languages, and audiences is not only more accurate and cost-effective but also significantly reduces the occurrence of fabricated information.
While the primary focus remains on the Indian market, some startups are also exploring international opportunities in regions such as the Middle East and Japan. For instance, Gnani’s voice bots have already found applications in Silicon Valley, assisting a California-based Harley-Davidson leasing company in reaching Spanish-speaking customers.
