Realtime Industrial Anomaly Detection
YOLO-based edge AI pipeline for Bar & Rod Mill video analytics - detects equipment anomalies and safety violations from live plant camera feeds with low-latency inference on edge hardware.
Building
I'm an M.Tech Mechatronics researcher at IIT Bhilai, working on real-time industrial anomaly detection with Edge AI, computer vision, YOLO-based video analytics and multi-modal RAG systems. Previously a Senior Software Engineer and AI R&D Engineer, building applied ML, robotics, OpenVINO/Jetson deployments and cloud-native systems across industrial, healthcare and education domains. IEEE volunteer, Intel Software Innovator and Arm Developer Ambassador.
Open to Computer Vision / Edge AI / Robotics roles
A snapshot of recent builds spanning edge computer vision, multi-modal LLM tooling and field robotics - all engineered for real deployment, not just notebooks.
YOLO-based edge AI pipeline for Bar & Rod Mill video analytics - detects equipment anomalies and safety violations from live plant camera feeds with low-latency inference on edge hardware.
LLM/RAG system that fuses sensor, vision and log data into plain-language plant intelligence alerts.
Framework that pairs static code analysis with semantic LLM feedback for fairer, faster grading.
NVIDIA Jetson Nano vision system paired with a 6-DOF parallel robot for targeted weed removal.
Exploration of zero-shot text-to-image generation pipelines and prompt-conditioning techniques.
Latest publication in ACM CODS 2025 - a framework that combines static code analysis with semantic LLM feedback to produce structured, rubric-aligned programming feedback across correctness, style and design.
Also: Adaptive Kalman-filter YOLOv5 tracking - SSRN Electronic Journal, 2022
Whether it's an edge-deployed vision model, a research collaboration, or a talk for your IEEE chapter - I'd love to hear about it.