Hamad Musa

Developed a Large Language Model (LLM) to support the DYEL Summer Program, part of Stanford’s SUPER (Summer Undergraduate Program on Energy Research). Designed the accompanying module infrastructure in Canvas, enabling participants to interact with LLM-powered resources and case studies. Contributed to Explore Energy’s cross-campus initiative to integrate AI into energy research education and outreach.

Contributed to the development of AI safety systems for micro-mobility platforms—including bikes, mopeds, and motorcycles—designed to prevent accidents through real-time computer vision and object detection. Tested and optimized end-to-end embedded vision pipelines running on edge devices validating performance under real-world conditions. This involved benchmarking ML-accelerated processors and stress-testing software. Collaborated cross-functionally with AI, hardware, and product teams to deliver highly optimized, deployable solutions that bridge safety-critical software with physical mobility systems.

Developed a deep learning pipeline that powered the company's core facial recognition system used to identify individuals in large-scale event photography. My work enabled fully automated, personalized photo delivery—scaling across thousands of attendees with high accuracy, even in challenging visual conditions. Designed and implemented a face matching system using facial embeddings, enhanced with image resolution techniques to improve performance in low-light, crowded scenes. Collaborated closely with a fast-moving startup team to deploy deep learning systems that delivered real-time inference and strong user satisfaction.

Selected for Stanford’s competitive CURIS research program, where I led the development of a deep learning model powering a breakthrough wrist-worn device for non-invasive, continuous blood pressure monitoring. Designed and fine-tuned a custom convolutional neural network (CNN) to predict blood pressure from real-time optical measurements of arterial diameter—transforming raw biomedical signals into actionable health insights. Collaborated with Stanford researchers under Dr. Parker Ruth to architect and optimize the full ML pipeline, from data preprocessing to deployment-ready inference.