Sanas Speech AI Joins Mavenir's MAVcore Voice AI Portfolio

Sanas is proud to bring our real-time Speech AI into Mavenir's MAVcore Voice AI portfolio, giving operators a new way to deliver voice AI natively from inside their own mobile core network.

Mavenir is the best-in-class telco giant behind AI-by-design networks that carry voice for more than 300 operators in over 120 countries, serving more than half the world's mobile subscribers. This relationship puts Speech AI in front of more carrier networks, and more callers, than almost any partnership could.

Mavenir's AI Agent Studio + Sanas then orchestrates and customizes the joint solution, giving operators service control, real-time performance, optimized operational costs, and the ability to build and monetize new conversational AI experiences.

Through this integration, Mavenir's MAVcore portfolio adds Sanas with:

• Real-time speech enhancement and HD Voice clarity
• Accent transformation and language understanding, so callers can be heard clearly in their preferred language
• Deepfake detection and fraud-signal capabilities, giving operators a carrier-grade way to catch synthetic-voice attacks

Why this belongs inside the network

Speech AI has mostly lived in the application layer, on top of the network rather than inside it. Mavenir and Sanas are changing that by embedding Speech AI directly within the operator's own network. It runs as sovereign AI, built to meet regulatory and data-residency requirements. That gives operators a way to turn voice from a utility they carry into a service they can differentiate, govern, and monetize. 

This isn't Sanas' first time operating at this level of stakes. Through our Language Assist partnership with MicroAutomation, Sanas already enables 911 dispatchers to hear callers in English as they speak, with no interpreter and no delay. Bringing that same real-time translation and speech enhancement into Mavenir's network extends it to a much larger share of the calls people make every day, at telecom scale rather than single-use-case scale.

Voice-cloning technology has also advanced quickly, and bad actors are already using it to commit fraud. That's part of why deepfake detection is built into this integration from day one: operators need a carrier-grade way to detect and defend against synthetic-voice attacks as that threat grows, not a bolt-on response after the fact.

"Voice remains the most trusted and universal service operators provide, and the next era is about making it intelligent," said Pardeep Kohli, President and CEO of Mavenir. "Together with Sanas, Mavenir is enabling operators to transform voice from a connectivity service into a monetizable, AI-powered experience. By embedding advanced Voice AI directly within the telecom network, operators can deliver real-time translation, AI-enhanced voice quality, and next-generation security while maintaining the performance, privacy, and sovereignty that modern communications require."

"Telecom operators sit at the center of some of the world's most important conversations," said Sharath Keshava Narayana, Co-Founder and CEO of Sanas. "Together with Mavenir, we're bringing a new layer of intelligence directly into carrier networks, creating new opportunities to improve communication quality, strengthen security, and deliver more value through every conversation. Communications networks have connected people for decades. We believe their next chapter is helping people understand one another more clearly."

Want to learn more? Join us together at the AI-Native Telco Forum, September 8 to 9 in Düsseldorf, and at FutureNet Asia, September 22 to 23 in Singapore.