Gnani.ai unveils voice-to-voice language model powering 10 Million calls per day with NVIDIA AI

New Delhi [India], October 24 : Gnani.ai has launched its speech-to-speech large language model (LLM) which can handle over 10 million voice interactions daily, the company said in a release on Thursday.

The LLM model is powered by the NVIDIA AI-accelerated computing platform, which includes NVIDIA AI Enterprise software for the development and deployment of generative AI applications.
The Gnani.ai model will transform the customer engagement and operational efficiency for over 150 leading enterprises across India and the United States, the company said.

Handling 10 million calls per day with ease, the model automates routine tasks, reduces operational expenses, minimises manual workloads, and empowers customer service teams to focus on complex, high-value interactions, as per the company.

The model was trained on multi-node NVIDIA Hopper GPUs using the NVIDIA NeMo end-to-end platform for developing custom generative AI--including LLMs, multimodal, vision, and speech AI--on over 14 million hours of proprietary multilingual conversational data supporting 14 languages, as per the company.
The model delivers faster, smarter, and more human-like responses--solving complex business challenges across industries, primarily in the banking, financial services and insurance (BFSI) sector along with other industries.
"We're at a pivotal moment in AI-driven customer service," said Ganesh Gopalan, CEO of Gnani.ai. "The speech-to-speech LLM is transforming the way enterprises handle their most critical customer interactions. By combining NVIDIA accelerated computing with our conversational AI expertise, we're helping businesses achieve scale and efficiency with a business impact of over USD 6 billion, and that's truly revolutionary."
Gnani.ai's LLM has already redefined the operational frameworks of over 150 enterprises across India and the United States.
Gnani.ai is a major player in conversational AI, specialising in advanced speech recognition, natural language processing, and voice automation technologies.