Why Rebuilding Telecom Infrastructure Is Critical For Next Wave Of Voice AI

India’s digital ecosystem is undergoing a massive shift as enterprises move from text-heavy apps to conversational voice interfaces. The commercial opportunity behind this transition is significant. According to industry estimates, the Indian voice AI market is projected to reach $1.8 Bn by 2030.

However, as businesses race to build voice-enabled products for the next 500 Mn non-English-speaking internet users, they are running into an often-overlooked operational challenge. While foundational AI models and speech-to-text technologies continue to expand rapidly, the underlying telecommunications infrastructure remains ill-suited to support automated voice intelligence at scale.

These infrastructure bottlenecks took centre stage during a panel discussion titled ‘How Companies Are Using Voice AI To Reach The Next 500 Mn Users’ at the Inc42 AI Summit 2026. 

Moderated by Inc42 senior editor Nikhil Subramaniam, the session featured Bolna founder Maitreya Wagh, Vobiz.ai cofounder and CEO Suman Gandham, and Cashfree Payments SVP and head of product Nitin Pulyani. The panellists discussed what it takes to deploy and scale voice AI systems in real-world environments.

Rebuilding The Outdated Telephony Pipeline

The industry’s current challenge is that advanced, real-time AI systems are being forced to operate on telecom routing systems built over a decade ago. While orchestration platforms like Bolna can seamlessly chain multilingual models together to sound human-like, the final execution frequently breaks down over the phone. 

Legacy telecom systems suffer from high latency, limited 8 kHz sampling rates, lack of automated noise cancellation, and rigid call controls, making them ill-equipped to support the demands of AI-powered voice applications at scale. 

Vobiz.ai’s Gandham highlighted this gap, arguing that infrastructure built for human-to-human communication cannot adequately support autonomous voice agents.

“The AI models change every day, but the telco layer is still sitting on old infrastructure made for human-to-human calls. Legacy players are retrofitting themselves for AI voice, but they are not building ground-up. You need an AI-first telco infra layer to handle deeper call controls, low latency, and structural noise suppression,” he emphasised. 

This infrastructure readiness dictates commercial success, especially for highly regulated sectors like fintech. Companies like Cashfree Payments see significant potential in deploying voice agents to replace cumbersome merchant onboarding paperwork and automate fraud verification calls. 

However, voice applications cannot scale if the underlying infrastructure struggles to handle traffic surges. The pace of growth is already evident. Vobiz.ai saw its call volume surge from 1 Lakh to 30 Lakh concurrent daily calls within a month and is now targeting 3 Cr daily calls by the end of the year. 

The panellists argued that the next wave of voice AI adoption will require more than just advances in language models. As enterprises look to serve millions of new Indian internet users through conversational interfaces, resilient, low-latency telephony infrastructure will be just as critical as the intelligence powering the applications. 

Ultimately, the winners will be companies that can build reliable, scalable infrastructure capable of supporting human-to-AI interactions in real-world environments.

The post Why Rebuilding Telecom Infrastructure Is Critical For Next Wave Of Voice AI appeared first on Inc42 Media.