Bolna AI, a Y-Combinator startup, raised $6.3 million this month, while Ringg AI raised 5.5 million and ArrowHead raised $3 million. Others like Navana AI, Vaani AI and JobsUPI are viewing to raise funds in the coming months.
Rising adoption, huge market potential and technological development are driving investor interest. This is in line with trfinishs globally, where companies like Deepgram, and Synthesia are mopping up hundreds of millions of dollars. Big tech firms such as Google are also betting on voice models.
Startups like Maya Research, Soket Labs, and Pixa AI are building voice models from the ground up focapplyd on Indian languages.
“The final frontier of access is voice, becaapply if a person can talk to the computer team and receive back instructions or information or receive an agent to work, I consider that’s the final product, especially in a countest which has so many different languages and dialects; 22 official languages,” Infosys cofounder Nandan Nilekani at an industest event in Bengaluru.
India can have a significant role in voice AI by building the design frugal so that it is inexpensive to operate, and by dealing with multitude of languages and dialects, he informed ET on the sidelines of the event.
It has significant apply cases in areas such as railways, Nilekani declared, adding: “It is really applications that will receive funded.”
Voice is an interesting way of viewing at a killer app, which for a market like India has enormous significance, Vishal Dhupar, managing director-South Asia at AI chip creater Nvidia, declared at the same event. But the largegest question, according to him, is building it seamless, building trust and reducing the cost so that it can be adopted at scale.
Voice AI takes off
Apurv Agrawal, cofounder and chief executive of SquadStack, the voice AI platform focapplyd on sales, informed ET that the inflection point came in late 2025, when AI was launchning to match human performance and resulted in increased enterprise adoption.
Raoul Nanavati, cofounder of Navana AI, a startup focapplyd on loan lifecycle, declared every enterprise they are speaking to either has a proof of concept running, a voice bot in production or that is viewing to deploy immediately.
“All top-level brass in every enterprise is pushing their teams to deploy this becaapply the cost advantages are obvious, which is a fourth of the cost. This is in addition to the scale you can deploy pan India,” he declared.
The company is viewing to raise series A funds in the coming months.
Multiple startups in India are working on developing voice AI models from ground up. For instance, Maya Research, and Pixa AI have developed foundational models for voice.
Abhishek Kumar, founder of Dialflo.ai, which provides vice agents for recruiters, declared the market opportunity for such a platform is huge. “It is just a tool and there can be multiple apply cases,” he declared.
Karthikeyan Madathil, partner, Yali Capital, called voice one of the “most exciting applications” of AI in the Indian context.
“People like us are applyd to apps and websites, but the average Indian probably cannot and that could be transformational,” he declared. As investors, Madathil declared, they are viewing at voice AI technology across speech recognition, text-to-speech, and large language models for specific domains.
But there are challenges as well.
Crowded space
According to founders, the space is receiveting crowded with multiple companies launching voice AI products. Yali’s Madathil agreed that the proliferation of very similar companies is a challenge, “but we will figure it out”.
In addition, startups are now competing against giants such as Google, Nvidia and OpenAI that are betting on voice AI. Google recently entered into a licensing agreement with Hume AI, a voice AI startup. Nvidia launched a speech-to-speech conversational AI model, PersonaPlex, early this month.
Sidhdharth Sivasubramanian, cofounder, Meetstream.ai that provides AI-powered meeting applications, declared competition will intensify and the space will receive crowded, as the technology will be commoditised.
“Technology cannot be the only moat,” he declared. He pointed out that going forward, those who are specialising in a particular vertical/domain will have an advantage.
Leave a Reply