What to Expect
As part of the Tesla Autopilot team, you will help lead the development of photorealistic, procedural environments that test, challenge and improve our Full Self Driving technology. Work alongside other artists, technical artists and simulation engineers to grow coverage of our virtual world to include more regions, variations, biomes and drivable topologies. 

Your goal will be to supercharge our end-to-end self driving neural networks by training them on automated procedurally created environments for anywhere in the world and beyond. The software, assets and virtual worlds that you create will convince the car it is driving in real locations to ensure the safety of our products by reproducing real driving behavior.

What You’ll Do
Use Houdini, Maya and in-house tools to create large, detailed road networks and associated assets in a procedural manner 
Utilise propriatary fleet data sources, neural network outputs and 3rd party spatial data to maximize location coverage 
Respond to data requirements and Full Self Driving product direction by providing realistic environments and assets 
Work closely with the AP Simulation team to continuously improve and automate asset creation workflows
What You’ll Bring
Portfolio and/or code examples demonstrating large-scale generation of procedural environments 
Excellent knowledge of Houdini, Python and Vex 
Extensive experience with automated 3D production pipelines and CI as used in film and game development 
Understanding of high performance assets in relation to ray traced rendering and graphics pipelines 
Pragmatic and able to deliver solutions to virtual world building problems with a first-principles approach 
Solid analytical & spatial math skills related to road topology and geometry generation
Experience with Git, GitHub Automation, Linux and Docker 
Proficiency with C++ and game/simulation engine programming 
Understanding of vision-based ML and experience with NN training

Palo Alto, California

Full time