Rivian is looking to develop, optimize, and deploy ultra-low latency Deep learning/Machine Learning algorithms for ADAS and Autonomy use cases to improve safety-critical self-driving features in their vehicles.
Requirements
- Good understanding of the fundamentals of deep learning, heterogenous computing, ML accelerators and compilers with 7+ years of industrial experience.
- Research and development experience in one or more of the following areas: model compression and neural architecture search techniques knowledge distillation, pruning, quantization and quantization aware training optimizing and deploying inference on various embedded processors
- optimizing and deploying inference on various embedded processors
- Experience defining compute architecture for efficient Deep learning inferencing.
- Capability to understand hardware spec documents and performance profiling tools.
- Strong Python programming background and in-depth knowledge of at least one framework amongst PyTorch, TensorFlow or MXNet
- Experience implementing inference logic from first principles using low level subroutines like BLAS, CUDA Kernels or C++ natively.
Responsibilities
- Develop, optimize and deploy ultra-low latency Deep learning/Machine Learning algorithms for Rivian ADAS and Autonomy use cases.
- Research state of the art model compression and efficient model design techniques and enable the team to leverage these across a wide range of customer facing features.
- Collaborate with the low-level software and hardware architecture teams to characterize the in-house ML models on our embedded platforms and optimize the models subject to the on device compute and memory constraints.
Other
- Good team player with great communication skills to drive cross functional efforts in a fast-paced development environment.
- MS. or Ph.D. in Computer Science, Electrical, Mechanical, Aerospace Engineering or a related field.
- 7+ years of industrial experience.
