OneRing is an intelligent wearable ring for Parkinson’s patients to better monitor movement disorder and prescribe medications. Launched product on Kickstarter. Successfully funded in 6 days, 240% funding of original goal, shipped product to worldwide backers (14 countries), 100+ daily PD patient users. Built iOS app, doctor web portal, machine learning diagnostic system, 3D printed rings - Full Stack. Internationally featured (Bloomberg Businessweek, FastCompany, NBC, Business Insider)

Top 10 paper presenter for machine learning research on early Schizophrenia diagnostics through diffusion MRI tractography analysis and computer vision. Worked under Dr. Lipeng Ning and Dr. Martha Shenton

Smartphone-based Diabetic Retinopathy (DR) diagnostic device through image processing and Random Forest machine learning. Developed a novel prognostic for DR by morphologically modeling hypertensive changes at arterial-venous crossings – therefore ultimately identifying new symptoms (venous compression, deflection, and tortuosity) to diagnose and monitor the condition. Project conducted under Dr. Rahul Khurana. Won International BioGENEius Challenge 2016.

Worked on longitudinal Diabetic Retinopathy study under Dr. Srinivas Sadda for automated retinal diagnostics. Modeled retinal blood vessels to quantify branch structures and identify relationships between retinal disease stages via automated processes.

Worked on GMCP-Tracker project in Dr. Silvio Saverese’s group under Dr. Amir Zamir to develop a Python algorithm for globally tracking multiple objects in video footage through generalized minimum clique graphs. Worked with CNN based Deformable Parts Models, OpenCV Haar cascade and HOG Classifiers, MATLAB video processing, and Python graph optimization. Deployed entire end-to-end tracker system on Stanford’s servers (http://cs.stanford.edu/~gmcp/).

Developed pupil tracker system and ML-based computer vision algorithms to detect abnormal saccades in Parkinson’s affected eyes. Worked under Dr. Rohit Dhall. Software and hardware went on to be utilized in Stanford research involving early PD diagnostics.