Building intelligent visual systems that see, understand, and act - from defect detection on factory floors to real-time safety alerts in warehouses.
Keypoint detection-based system for automatic reading of analogue pressure gauges from industrial camera feeds. Handles reflections, uneven lighting, and dial variations with a custom dataset and inference pipeline for real-time extraction.
Computer vision safety system detecting proximity between forklifts and workers in warehouse environments. Multi-object detection with real-time distance estimation logic and fine-tuned models for domain-specific video data.
Real-time chess board state detection using YOLOv8, TensorFlow, and OpenCV. Full pipeline for image capture, inference, and state reconstruction with reliable detection across multiple board designs and camera angles.
Series of computer vision POCs for manufacturing quality control - glass bottle inspection, biscuit dimension measurement, and packaging integrity checks. Demonstrating automation potential across critical quality workflows in FMCG and industrial settings.
I'm a Computer Vision and Gen AI Engineer based in Hyderabad, India. I specialize in building production-ready visual intelligence systems that work in real-world, often harsh conditions - factory floors, warehouses, industrial plants.
My work spans the full stack of CV engineering: from dataset curation and model training to edge deployment with TensorRT and DeepStream. I care deeply about models that are fast, accurate, and actually usable in production.
Beyond the technical side, I enjoy the client-facing dimension of this work - translating ambiguous problems into clear solutions, and building POCs that earn trust.