Mixed Reality
Built synthetic pipelines (C++) to simulate devices and generate synthetic sensor data (IMU, images, etc) for multiple types of Mixed Reality products (controllers, headsets, etc.). The synthetics pipelines were used by different teams across Mixed Reality for hardware prototyping, AI and CV algorithms development. Synthetics data were created using hyper realistic environments for targeted products and scenarios, which dramatically improved the hardware and software iteration speed.
Conceived and implemented the simulation workflows (C++, Python, C#) of synthetic human (SMPL) with virtual devices for environment understanding teams. Added the new capabilities to generate simulation data with animated humans for complex multi-player, multi-device scenarios.
Designed and implemented micro services (Python) leveraging Nvidia Omniverse to provide a synthetic platform (USD) for engineers to compose synthetic scenes, prototype virtual devices, generate and visualize trajectories, ect., in real time.
Built metrics pipelines (C#) to evaluate overall product quality through Ground Truth data with a massively distributed computational and storage platform.
Implemented telemetry pipeline, analyzed in-market telemetry data, and built data visualizations with Power BI to quantify customer focused product performance and quality issues, resulting in software updates and hardware design changes.
