Software Engineering Leadership in Reality Labs:

• DSLs / compilers / low level optimization (AVX512, AVX2, NEON, Compute Shaders, OpenCL, CUDA)

• Low latency, realtime GFX / ML pipelines,

• ML inference on devices (headsets, et al.),

• VR/MR Super Resolution (temporal super resolution + ML artifact mitigation),

• Architecture for large scale graphics + ML workloads.

• Hired as an external IC7 and survived ;)

• Sprawl - a compute graph framework for large scale heterogeneous computation,

• Machine Learning for robotics, games & computer vision,

• Hybrid rasterization / raytracing pipelines.

Built a simplistic, lightweight, scalable raytracer for ML/RL workloads, around a custom Python DSL -> CUDA / AVX512 compiler/transpiler. Acquired in June 2020 by Unity.

• Teaching metal appendages how to manipulate a Rubik's cube (OpenAI Dactyl),

• ORRB - OpenAI Remote Rendering Backend,

• Simulation,

• Rendering,

• Computer Vision,

• Low Level Performance Optimization.

Senior Software Engineer in Brain Robotics

• Generating rich 3d environments for computer vision / RL

Senior Software Engineer in Google Brain (Google Research & Machine Intelligence)

• Matrix multiplication done fast.

• Automatic optimization and compilation of numerical methods

• Operation fusion and auto vectorization

• Mobile TensorFlow