Design scalable ML workflows for generating AV scenarios - ranging from traffic simulations to synthetic scenario creation from a natural language specification.
Contribute to tooling for migrating scenario creation using the traditional approach to AI-based scenario understanding and validation
Own the development and upkeep of the simulation framework, driving its reliability and performance to unblock all technical stakeholders.
Collaborate directly with internal customers and partner teams to provide solutions for their test creation workflows.
Have an opportunity to tackle simulation and autonomy evaluation challenges across various areas of the autonomy system, including motion planning, mapping, Controls.
Create and extend software for behavior, pathfinding, road networks, spatial queries, collision detection, vehicle control, vehicle dynamics, etc.
MS or higher degree in Computer Science, Machine Learning, or related field
Good development skills in modern C++ and Python
Proficiency in ML libraries and experience in transformer and diffusion architectures
Basic understanding of generative models
2+ years of Industry experience in ML and C++ software
Familiarity with AV simulation environments and driving datasets
As a Software Engineer on the Vehicle Motion Planning & Control team, you will design algorithms and implement software to enhance the autonomous driving capabilities of vehicles.
As a Software Engineer on the Identity Infrastructure Engineering team, you'll create and operate security tools and infrastructure for multi-cloud environments.
Build and ship end-to-end features across gateway infrastructure, enterprise deployments, and agentic systems, making architecture, product, and security decisions while working closely with customers and AI agents.
Develop and maintain software that connects aircraft avionics, simulation models, test hardware, and engineering tools, enabling hardware-in-the-loop testing and helping teams diagnose integration issues before flight.
Extend the compiler frontend and intermediate representation to bring up emerging LLM architectures, preserve their quantization and numerical behavior, and validate models on simulators and FPGA prototypes.
