• Designed early-stage research pipeline in collaboration with Stanford Medicine faculty to explore multimodal clinical prediction in low-resource hospital settings.
• Investigating robust generalization, fairness, and bias mitigation for real-world deployment.
• Prototyping visual grounding metrics to evaluate model stability, hallucination patterns, and clinical interpretability.
• Drafting a research direction focused on scalable ML for under-resourced environments, aligning with Stanford's translational AI principles
• Built a monocular-metric depth estimation pipeline using ROS1, Ubuntu, and Intel RealSense for robotics navigation.
• Developed preprocessing, calibration, and evaluation tools to improve metric depth prediction accuracy.
• Conducted benchmarking experiments to analyze system performance and identify weaknesses in perception models.
• Transitioned into Visual Grounding research, investigating hallucination patterns across models like GroundingDINO, GLIP, and OWL-ViT.
• Designed and developed a complete multi-page website (Home, Services, Hours, Location, Contact) for a local auto repair business.
• Implemented responsive layouts, improved user flow, and created a modern interface tailored to client needs.
• Deployed using GitHub + Vercel with version control, ensuring consistent updates and high availability.
• Completed as a community service project to enhance visibility and digital presence for a local family business.
• Designing a social platform focused on neighborhood-level communication, safety alerts, and real-time local engagement.
• Prototyping scalable backend features including authentication, messaging, and geospatial event discovery.
• Conducting user and market research to validate core features and refine platform usability and retention.
• Early product engineering initiative under your startup O1Theta, focusing on community-driven technology impact.
