Tata Elxsi and Syntiant to build AI Voice Apps

Tata Elxsi, a design and technology service provider company, has now announced a partnership with Syntiant to develop AI-powered voice applications.

Syntiant is founded by Jeremy Holleman in the year 2017. It helps firms to move AI and Machine Learning from the cloud to edge devices.

Speaking about this partnership Nitin Pai, Chief Marketing and Strategy Officer, Tata Elxsi, said, "Tata Elxsi with its Artificial Intelligence Centre of Excellence addresses the growing needs of intelligent systems. We work on various AI and machine learning areas. The problem statement varies from use-case where AI can enhance everyday life quality through intelligent smart appliances to enable predictive analysis, anomaly detection in a factory environment, and advanced autonomous driving. By intersecting design thinking with advanced technologies, we help our clients bring innovative AI-enabled edge devices to the global market using Syntiant AI processors."

Mallik Moturi, Vice President of product and business development at Syntiant, said, "Syntiant is in the business of deep learning and semiconductor design to advance machine learning capabilities in edge devices. Our neural decision processors provide superior performance at a fraction of the power of any DSP on the market today. We're collaborating with Tata Elxsi to combine our resources and make it easy for customers to embed AI in almost any smart device, from wearables to TV and key fob remotes to set-top boxes and beyond, with minimal or no drain on battery power consumption."

As a part of the deal, the firms will offer end to end solutions and services to help Syntiant’s customers to develop low power, always-on voice applications for products across multiple categories, which include consumer electronics, smart home devices as well as automotive and industrial use cases.

With this new partnership with Tata Elxsi, the manufacturers using the NDP Processors in their AI devices would get a turnkey solution that brings high accurate and low latency inference to enable several near and far-field voice applications, which includes event detection, multiple wake words, keyword spotting, local commands recognition, and speaker identification.