CS graduate from BUET with a published research background in Vision Transformers (MIDL 2026) and hands-on experience in on-device model optimization, edge inference, and PyTorch workflows. Proven track record in benchmarking multi-architecture models (XGBoost, 1D-CNNs, ONNX) for sub-millisecond execution on physical devices. Seeking an opportunity to contribute in next-generation efficient AI and vision systems.