Maya Research Raises $1.9 Million Seed Round Funding from South Park Commons to Build Voice Interface for the world

• Maya is the only Indian company on Speech Arena with #6 Open Weights Ranking
• 440,000+ model downloads on Hugging Face from developers globally 
• 3 million+ app downloads across India, Southeast Asia, and MENA 

Maya Research, an AI company building a voice interface for the Majority of the World, has raised $1.9 million in a seed round led by South Park Commons. 

Maya is building a voice interface for a world where billions of people interact with technology through conversation rather than typing, tapping, and navigating complex interfaces. At the core of this interface are conversational models designed to speak, think, and respond like a native speaker, adapting to local languages, contexts, and conversational nuances. While most AI products still treat voice as a microphone attached to a text system, Maya is building conversational intelligence that enables more natural interactions between people and technology. 

"India is not a text economy, it never was. While the conversation around AI sovereignty has focused on large language models and text interfaces, the more urgent question is: whose voice models are Indians talking to? Right now, that answer is almost entirely foreign. Maya exists to change that. We built our models frugally, on cloud credits, because we believed the problem was worth solving regardless of resources. And what we have built is already on the leaderboard. For the next five billion people, voice is not a feature. It is how they live. They will access AI through an interface that can hold a real conversation with them and understand them. That is what we are building. We want that experience to be built in India, owned in India, and available to the world from India," Dheemanth Reddy, Co-founder and CEO of Maya Research. 

Founded in 2025 by BS Dheemanth Reddy and Bharath Kumar Kakumani, Maya Research has crossed 440,000 model downloads on Hugging Face, amassed 3 million app downloads across India, Southeast Asia, and MENA, and become the only Indian company on Speech Arena, where its open-source voice model Maya 1 ranks among the leading open-weight models globally. 

Maya 1, the company's open-source emotive voice model released under Apache 2.0, holds a Quality Elo of 1,051 on Speech Arena, placing it 6th among open-weight models globally and on par with OpenAI's GPT-Realtime-2. While Maya's consumer application creates a data flywheel for high-quality, hard-to-collect data, it provides a unique advantage for the company to create models faster and better, continuously. 

"Technology should feel magical to people. It should feel like something that belongs to them, not something foreign they have to learn. For too long, technology has alienated the majority of the world because it was never built for how they speak, think, or live. We are building models that talk like one of them, understand like one of them, and feel local from the first interaction. When technology feels native, real adoption happens," said Bharath Kumar Kakumani, Co-founder and CTO of Maya Research. 

"The internet was built around English and text, which quietly left most of the world outside the interface. Maya is rebuilding that interface around how people actually communicate: by speaking, in their own language. As voice becomes the way billions of people interact with technology, the company with the richest multilingual speech data and the strongest conversational models will have a defining advantage. Maya has already shown strong early proof, with over 3 million app downloads and more than 440,000 model downloads on Hugging Face," said Prateek Mehta, General Partner at South Park Commons. 

“AI for India will require re-imagining both the input modalities and the interfaces through which our population will access this life-changing technology. Maya is India’s ChatGPT moment,” said Aditya Agarwal, General Partner at South Park Commons 

While voice AI has made significant progress in transcription, translation, and speech generation, the challenge of building systems that can participate in natural human conversation remains largely unsolved. For Maya's founders, who grew up in the small towns of Vempalli and Ongole in Andhra Pradesh before studying at New York University, that challenge represents one of the largest opportunities in technology. As billions of people come online through voice rather than keyboards, the company believes the next breakthrough will come from systems that can listen, understand, and respond as naturally as people do.