Bishmoy Paul
Graduate Research Assistant at Santa Clara University, Competition Master at Kaggle
Hi, I’m Bishmoy Paul, a Graduate Research Assistant at Santa Clara University, currently focusing on optimizing Large Language Model (LLM) Inference using Quantization and Sparsification methods implemented with custom CUDA kernels. My experience spans applying deep learning across diverse domains, including synthetic image and song detection, video-text retrieval, and Retrieval-Augmented Generation systems. These projects have equipped me with hands-on expertise in leveraging cutting-edge technologies like multimodal transformers and deep learning frameworks such as PyTorch and TensorFlow.
I’ve also earned top placements in IEEE competitions and Kaggle challenges, where I demonstrated my ability to solve complex, real-world problems with innovative solutions. My goal is to bridge cutting-edge research with real-world solutions in AI and machine learning.