Bengaluru- and Singapore-based Vaani AI Research has opened its voice AI stack to the wider developer community with a self-serve platform for building, deploying and managing voice agents. The company says the same orchestration layer that already runs voice agents for its enterprise clients is now available to developers, startups and product teams, who no longer need to stitch together speech models, telephony and conversation infrastructure on their own.

From description to a live voice agent

On the new platform, teams can describe the agent they want, preview how it sounds, configure the actions it can take and then deploy it across phone calls, websites, mobile apps and WhatsApp. Builders keep control over how the agent responds, how it holds on to conversational context, which tools it can call and when it should hand a conversation over to a human.

Agents can be built either through APIs or through a configurable no-code studio. According to the company, the studio includes a visual flow builder, prompt management, call review and evaluations, and supports 22 Indian languages alongside English. Vaani lists developer access to its stack at ₹5 per minute, all-inclusive.

The ‘voice harness’ at the core

The centrepiece of the launch is what Vaani calls its voice harness, an infrastructure layer that links the company’s own speech models with everything a voice agent needs to work in production. It pulls together inference optimisation, memory and context management, conversational orchestration, telephony, tool integrations and human-in-the-loop workflows into a single stack.

The pitch targets a familiar gap in voice AI: agents that perform well in a demo but stumble in real conversations, where callers switch languages mid-sentence, interrupt, speak over background noise or say something unexpected. Vaani says its platform lets teams refine these interactions without leaning only on heavy prompt engineering.

“Seventy-five percent of the world does not primarily speak English, and most voice AI still treats that as an edge case. We built this harness inside live enterprise deployments, where every call shows you what breaks. Today we are opening it to anyone who wants to build a voice agent in the language their customers actually speak,” said Tushar Shinde, Co-founder and CEO, Vaani AI Research.

Self-serve for builders, engineers for enterprises

The platform grew out of Vaani’s enterprise deployments, where the company built its own speech models, inference infrastructure and orchestration layer to govern how agents listen, respond, act and remember. Its infrastructure is already in use by businesses in India and overseas, with the company citing customers across banking, lending, insurance, real estate, healthcare and logistics.

The self-serve route sits alongside a white-glove option: enterprises with specialised needs can still work with Vaani’s forward-deployed engineering team on integrations and workflow optimisation.

Founded in 2025 by Tushar Shinde, Nitesh Tripathi and Nitish Mishra, Vaani AI Research is backed by Venture Catalysts, the Google AI Accelerator and an undisclosed institutional investor.

Why it matters

Voice is fast becoming the main interface for AI in India, where a large share of new digital users are more comfortable speaking than typing, and in their own language. By packaging speech models, telephony and orchestration into one developer-facing product, Vaani joins a growing set of Indian voice AI players trying to make multilingual, production-grade voice agents a few clicks away rather than a months-long integration project.

Be a part of Elets Collaborative Initiatives. Join Us for Upcoming Events and explore business opportunities. Like us on Facebook , connect with us on LinkedIn and follow us on Twitter.

"Exciting news! Elets technomedia is now on WhatsApp Channels Subscribe today by clicking the link and stay updated with the latest insights!" Click here!