Chennai: embedUR, embedded systems and edge AI company, expects its edge artificial intelligence business overtake its services business, driven by industrial demand for deploying AI systems on-site. It also deepens its partnership with global chip vendors. The US-headquartered company, which builds firmware and software solutions for IoT devices and networking applications, has pivoted to edge AI business in recent years, with work done by its Chennai team. Rajesh Subramaniam, founder and CEO of embedUR systems, said they expect the demand from industrial applications in medical devices, logistics and transportation, and commercial lighting to adopt the edge AI models for low-latency workloads.
It built a platform called ModelNova, which offers a library of pre-trained AI models for specific applications. It aims to eliminate the need to use proprietary toolkits from individual chip vendors. Along with existing partnerships with global chip design vendors such as Infineon, ST Micron, Synaptics, it has expanded with Arm, Silicon Labs, NXP Semiconductors, and Ceva. The chips capable of hosting AI models would be taken up by original equipment manufacturing companies. “The library will make it easy for any software developer to be able to start creating and customising models with the point and click interface,” he said. He said their edge AI platform revenue, which currently stands at 10%, would overtake their primary business of embedded systems’ revenue by 2027. Subramaniam said the company plans to spin off Model Nova as a fully owned subsidiary before year-end, separating the product business from services. The company aims to complete Rs 300 crore investment in its Chennai centre as part of its Rs 500 crore commitment.
