Gaurika Nawani

I’m a third-year B.Tech student specializing in Artificial Intelligence and Data Science, with hands-on experience turning models into end-to-end systems. In my Data Science Internship at Quick Heal, I engineered and deployed a production-ready malware detection pipeline using a custom computer vision model, processing 200,000+ files (~16 GB) in a live environment, and analyzed security tool data inside VM environments to improve threat detection.

I’m especially proud of my accuracy-focused work: I built a speech emotion detection system that achieved 97.25% accuracy, and I developed an ISL fingerspelling recognition approach with ResNet50 (99.06% accuracy) plus LSTM sentiment analysis (97.64% accuracy). Across projects, I use Python and deep learning with libraries like PyTorch/TensorFlow, and I focus on explainability and real-world robustness—supported by tools such as SHAP, Librosa, and OpenCV.