Shreyas Chitransh

I am a Vellore Institute of Technology graduate with a passion for building intelligent systems that bridge the gap between raw data and actionable insights. During my studies, I designed a real-time Traffic Sign Recognition model using Python, OpenCV, and Keras—training a convolutional neural network on diverse datasets to achieve over 90% accuracy under varying lighting and orientation conditions. I also developed a Gesture-Based Virtual Mouse leveraging Mediapipe’s hand-tracking capabilities, translating natural hand movements into system commands and refining gesture classification algorithms to enhance interaction efficiency. In my current role at Nference, I have been instrumental in advancing our AI-driven diagnostic assistant by designing and deploying a conversational interface that enables clinicians to query ECG data, optimizing backend pipelines for rapid model inference, and containerizing machine learning components with Docker to ensure seamless deployment and scalability. Throughout these experiences, I have focused on model development, data augmentation, and inference optimization—always striving to transform complex signals into meaningful, real-time predictions.