SwitchOn Inc has closed an $8 million Pre Series B funding round, with participation from IvyCap Ventures, SIG Tattva, the corporate venture arm of Somany Impresa Group, and Trifecta Capital. The company builds an AI powered visual inspection system called DeepInspect, designed to help factories spot product defects that are nearly impossible for human eyes to catch.
According to the company, the core challenge it set out to solve is not about training people better, but about the limits of human perception itself. On a fast moving production line, where a bottle can pass a checkpoint at 400 units per minute, a worker is expected to identify a crack finer than a strand of hair. SwitchOn says this is essentially a physics constraint rather than a skills gap. Its DeepInspect system claims to detect such defects with precision down to 150 microns and an accuracy rate of 99.9 percent.
Traditionally, setting up a new inspection system for a product line has taken factories several weeks. SwitchOn says its platform cuts that down dramatically, allowing a new product to go live in around 45 minutes using roughly 200 images, without requiring any actual defective samples for training.
The company points to real results from its deployments. One bottling plant that adopted DeepInspect reportedly brought customer complaints down to zero while reducing its overall cost of quality by 80 percent. SwitchOn also states that its system has been used to inspect more than 1,000 different products, ranging from bottles and tablets to connecting rods, motherboards, and soap bars, essentially anything that moves along a production line.
Scaling Across Continents
DeepInspect is currently active on more than 170 production lines across over 60 factories, spread across four continents, according to the company. SwitchOn describes this growth as part of a larger shift in artificial intelligence, moving beyond software and screens into physical manufacturing environments. The company says it is building this technology from India for global manufacturers, and is currently hiring across most roles as it expands what it calls its work in physical AI.
