Nitish Rathore

Background

About Nitish Rathore

I am a software engineer passionate about Computer Vision and systems. My experience spans from deploying deep learning models for computer vision tasks to architecting scalable backend infrastructure for AI-driven solutions. At Pronisi, I played a foundational role in working cross teams and designing real-time AI video processing pipelines,(YOLO-V8, MediaPipe, Detectron-2, ResNet) for computer vision applications like fast-moving object detection and human pose estimation, achieving over 80% accuracy. Managed end-to-end data pipelines, automating data ingestion, preprocessing, and model deployment on AWS, Ray Serve. During my time at I worked on a research project focused on eliminating noise in multispectral images using deep learning to generate Land Use and Land Cover (LULC) maps. This research evolved into my capstone project at PES University, where I extended the work by integrating spatio-temporal analysis for urban growth modeling. Our findings were later presented at a conference in China, where I was awarded Best Presenter. I am passionate about solving challenging engineering problems, optimizing large-scale AI/ML systems, and continuously learning to push the boundaries of real-time intelligence and automation. Looking forward to collaborating on impactful projects that shape the future of AI-driven infrastructure.