Gustavo Braga Email & Phone Number

State Of São Paulo, Brazil

Nova York, Estados Unidos

Doge Labs is building Big Ballers VR (BB) for Meta Quest. BB has tens of thousands active players and stands in the top 15 apps on the Meta store. I'm tasked with developing new game modes, optimizing and writing shaders/VFX for the game.

São Paulo, Brasil

As an undergraduate assistant professor, I teach and mentor students in advanced topics in computer graphics and game development, such as shaders, ray tracing, ray matching, optimization techniques, and render pipelines.

São Paulo, Brasil

- A member of the R&D team tasked with researching, optimizing and integrating cutting-edge technologies into the realm of AR/VR. Built and deployed applications across multiple devices, such as Hololens, Oculus Quest, Vision Pro, mobile and Magic Leap. Those applications involved utilizing advanced technologies like LiDAR for point cloud data scans and manipulation, as well as Spatial anchors, Machine learning, and Digital Twins. To achieve this, I frequently had to develop efficient solutions using shaders, multithreading, and GPU programming. These solutions were designed to handle complex systems that integrated computer graphics, machine learning, and TCP/IP or UDP connections between devices. I gained experience working with various frameworks such as 8th wall, babylon.js, three.js, AFrame, and Unity, which were instrumental in developing AR/VR applications.- Engineered and optimized a specialized GPU and CPU multithreaded CODEC, akin to JPEG, employing HLSL for cloud rendering within Unity, delivering over 60fps with and enabling up to 300x compression per frame.- Programmed a high-performance CPU-based AR point cloud scanning app for mobile phones equipped with LiDAR for the broker market. Achieved +60fps performance and scanned over 1.620.000 points processed per second locally.- Assembled a real-time photorealistic digital twins app for retail, connecting and streaming frames from multiple cameras and scenarios to an AI system for product labeling. It enabled virtual training and seamless physical testing,eliminating manual labeling and scenario recording for substantial time and effort savings.

São Paulo, Brasil