Chacko Martin Koyikka

Chacko is a computer vision engineer and researcher specialized in low-latency convolutional neural network (CNN) deployment on constrained edge microprocessors. Over the past 9 years, he has designed embedded machine learning routines for logistics tracking and automated sorting systems, focusing on low-power devices.

"Computer vision algorithms shouldn't be locked away in theoretical cloud instances. The real value is realized when you drop the model down directly next to a physical conveyor belt to actuate instant pneumatic rejections. I co-founded ARCAISYS in May 2026 to prove that you can build extremely robust visual classifiers that perform reliably on-site in noisy, dusty, thermally unpredictable manufacturing settings."

At ARCAISYS, Chacko leads the engineering roadmap for the SightQC inspection kernel. He architected the edge inference pipeline that runs YOLO models inside Docker containers on local AWS IoT Greengrass boards, maintaining sub-100ms processing cycles to drive real-time solenoid rejection gates.