• Led design and implementation of a novel neural decoding architecture for the N1 implant, incorporating real-time intent prediction and adaptive feedback mechanisms, reducing error rates in complex UI interactions (e.g., cursor navigation, text entry) by 30%.

• Developed closed-loop patient interaction systems with optimized visual and predictive feedback, delivering natural, low-latency control experiences comparable to native motor function.

• Built high-bandwidth, secure data pipelines processing thousands of neural channels with sub-millisecond precision, adhering to FDA and IEC 62304 standards for safety-critical medical software.

• Collaborated closely with clinical teams and patients to iterate on application features, improving calibration workflows and long-term usability.