Zoho invests USD $10m in NVIDIA AI for language models

Zoho has announced a strategic decision to leverage NVIDIA's AI accelerated computing platform to enhance the development and deployment of its large language models (LLMs) within its SaaS applications.

The technology company plans to use NVIDIA NeMo, part of the NVIDIA AI Enterprise software, in the deployment process. This development is set to benefit over 700,000 Zoho customers around the world. Zoho has stated that it invested more than USD $10 million (AUD $14.9 million) in NVIDIA's AI technology and GPUs over the past year and intends to invest an equal amount in the coming year.

"Many LLMs on the market today are designed for consumer use, offering limited value for businesses," said Ramprakash Ramamoorthy, Director of AI at Zoho Corporation. "At Zoho, our mission is to develop LLMs tailored specifically for a wide range of business use cases. Owning our entire tech stack, with products spanning various business functions, allows us to integrate the essential element that makes AI truly effective: context."

Zoho places a high priority on user privacy in its AI model development to ensure compliance with privacy regulations, enhancing the models from inception rather than making adjustments later. The company aims to assist businesses in achieving a rapid return on investment by utilising NVIDIA's AI technology to increase operational speed and reduce delays.

With a decade-long experience in building AI technology, Zoho integrates AI within its extensive portfolio of over 100 products under its ManageEngine and Zoho brands. The AI approach is designed to be multi-modal, focusing on deriving contextual intelligence to aid business decision-making. By developing a range of narrow, small to medium-sized language models, distinct from LLMs, Zoho offers varied model sizes tailored to specific business needs, facilitating AI usage even for businesses with limited data.

"The ability to choose from a range of AI model sizes empowers businesses to tailor their AI solutions precisely to their needs, balancing performance with cost-effectiveness," explained Vishal Dhupar, Managing Director, Asia South at NVIDIA. "With NVIDIA's AI software and accelerated computing platform, Zoho is building a broad range of models to help serve the diverse needs of its business customers."

This collaboration enables Zoho to expedite the deployment of its LLMs using NVIDIA accelerated computing platforms, including NVIDIA Hopper GPUs and provides access to NVIDIA NeMo's end-to-end AI development platform. Zoho is also testing NVIDIA TensorRT-LLM to optimise LLM deployment, noting a significant 60% increase in throughput and a 35% reduction in latency compared to previous frameworks. Moreover, Zoho is accelerating additional tasks such as speech-to-text processing using NVIDIA's computing infrastructure.