Software Engineer, but More Expensive(TM).

On the Platforms Infrastructure team, building robust and scalable distributed infrastructure on which all of glean runs

Led a Brain-Robot-Interface project through the computational neuroscience phase, iterating on deep and classical statistical methods to classify EEG data collected from human subjects imagining movement (motor imagery) with high accuracy (45% -> 85%, 4-way) allowing humans to interface with robots to perform general day-to-day tasks (e.g. household chores). Under the supervision of Ruohan Zhang and Feifei-Li. CoRL 2023.

• Built the company's first AI-generated audio detector: https://docs.thehive.ai/docs/ai-generated-audio-detection

• Distributed training of large multi-head image classifiers to perform at super-high accuracy

• Built data sourcing / filtering tools (embedding / vector search, automatic image captioning, efficient sampling, etc.)

Using deep learning to model EEG data to predict various types of pathological behavior for the goal of developing therapies for such pathologies.