Dhruv Gautam

At Wipro, through the Teachnook & IIT Roorkee Program, I worked as an Artificial Intelligence Intern on supervised machine learning pipelines using Python, TensorFlow, Scikit-learn, and PostgreSQL. I built automated feature extraction workflows and data-cleaning queries, contributing to a 12% boost in predictive accuracy.

For my Smart Crop Weed Identification project, I trained a custom CNN transfer-learning model on agricultural images, reaching 95% validation accuracy and an F1-score of 0.94. I also developed image preprocessing and deployed the model for mobile edge inference with Dart and Flutter.

I built an AI Fitness Trainer project using MediaPipe Pose and OpenCV to process workout video streams and track repetitions. Its state-machine thresholding achieved 86% counting accuracy under multi-angle lighting conditions.

I'm studying Computer Science & Engineering at Bennett University and have solved 385+ data structures and algorithms problems in Java and C++. My work and coursework include computer vision, deep learning, and generative AI.