San Francisco, California, United States
Engineer with 6+ years of experience developing software and machine learning (ML) approaches for brain-computer interfaces (BCI). Co-developed a domain adaptation approach for calibration-free stabilization of neural signals. Created open-source real-time software for BCI experiments that has been adopted by leading clinical trials and lowered the barriers to deploying ML models in low-latency BCIs. Seeking roles in machine learning and software engineering
Member of the Systems Neural Engineering Lab (PI: Chethan Pandarinath, PhD) - Developed a real-time brain-computer interface software framework achieving sub-millisecond communication latency with native ML library support, adopted by multiple leading clinical trials to replace legacy platforms - Optimized and validated an unsupervised deep learning pipeline for calibration-free neural signal stabilization, demonstrating effectiveness on non-human primate datasets and establishing feasibility for real-time implementation - Managed end-to-end infrastructure operations including GPU servers, file servers, and real-time signal processing systems, ensuring 99%+ uptime for critical research operations - Collaborated cross-functionally with neurosurgeons, neurologists, and scientists to launch a clinical trial site and advance research directions toward clinical translation
- Developed and deployed real-time deep learning algorithms for neural signal denoising - Supported clinical research operations across 8 intracranial electrode recording sessions with epilepsy patients
- Re-engineered server-based haptic robot control system into cost-effective embedded solution using Simulink and Raspberry Pi, reducing costs while preserving a 500 Hz closed-loop control rate - Built full-stack Django web application managing large-scale automated rodent training system with MySQL integration, implementing adaptive difficulty algorithms that improved training efficiency - Maintained and troubleshot complex experimental setups, ensuring consistent data collection across multi-month research studies Skills practiced: - Programming (MATLAB, Python, Linux kernel customization, Bash scripting, Simulink) - Hardware design (Raspberry Pi, Arduino) - Animal care (rodent survival surgery, behavioral training) - Project management (resource allocation, Gantt charts, agile software development)
Graduate Teaching Assistant for BMED 3110 Quantitative Engineering Physiology Laboratory I - Led weekly 30-person lab sessions in which students were taught to build circuits for electrocardiography and electromyography - Taught the fundamentals of data acquisition and signal processing using Matlab, Python, and National Instruments systems - Held both remote and in-person office hours to accommodate social distancing requirements during the COVID-19 pandemic
Researched hardware and software implementations of voice sampling in patient care settings with the goal of making it easy for doctors and patients to use NeuroLex's vocal analysis as a diagnostic tool for neurological disorders
- 80 hours of volunteering in a post-anesthesia care unit at Emory Johns Creek Hospital - Observed three orthopedic and two GI surgeries - Shadowed anesthesiologists to learn about patient sedation and pain management - Learned best practices for bedside care from nurses attending to recovering patients