driven by the challenge of designing AI systems for real-world problems—whether in computer vision, NLP, or TSF. My work covers the full lifecycle: building, training, and evaluating models with modern frameworks, all while practising CRISP-DM like methodologies.
Research-wise, I am exposed to various topics. Currently persuing my M.Sc. with focus on Knowledge Distillation and Symbolic Reasoning, I do sometimes read articles on computer vision, machine unlearning and federated learning. I am especially interested to apply AI in healthcare and medicine.
I remain open to both industry-focused AI/ML roles and also academic R&D roles.
I rely on these tools primarily for data cleaning, EDA, and feature engineering—transforming raw data into clear, actionable insights and preparing it for seamless integration into ML/DL pipelines.
Started with TF&Keras but now I am mostly using Torch to design, train and evaluate ML/DL frameworks.
These libraries form the backbone of my computer vision workflows. I use them extensively to fine-tune SOTA models—such as YOLO varients for object detection and Mediapipe for pose estimation—on custom datasets.
