Software Engineer, Machine Learning

Overland AI•Seattle, WA

98d•$120,000 - $300,000•Remote

About The Position

We are looking for machine learning experts to join our team to advance the state of off-road autonomy. Experienced, well-qualified remote candidates able to work during Pacific Time Zone business hours will be considered. In this role, you will design, prototype, and deploy machine learning algorithms related to perception, state estimation, and behavior planning (Python). You will identify, implement, and respond to meaningful metrics that affect vehicle behavior. You will serve as a technical expert for the team balancing short-term deliverables and long-term directions. You will collaborate closely with experts across disciplines to work across the entire stack to improve the reliability and performance of the system.

Requirements

MS/PhD with publications in top-tier ML/CV venues (ICML, NeurIPS, ICLR, CVPR, ICCV, ECCV), OR equivalent years of industry experience in an applied research setting working on multi-year problems.
Expert in a relevant ML topic such as sequential prediction, structured prediction, inverse reinforcement learning, transformers, deep learning optimization, etc.
Excellent linear algebra, probability theory, and optimization knowledge.
Extensive experience in Python and a deep learning framework such as Pytorch, JAX, Tensorflow, etc.
You excel in a small-team atmosphere, taking ownership of problems and working with your colleagues to solve problems across disciplines.
Ability to obtain and maintain a DOD Security Clearance.

Nice To Haves

Knowledgeable in any related technologies such as C++, CUDA, ROS, etc.

Responsibilities

Design, prototype, and deploy machine learning algorithms related to perception, state estimation, and behavior planning (Python).
Identify, implement, and respond to meaningful metrics that affect vehicle behavior.
Serve as a technical expert for the team balancing short-term deliverables and long-term directions.
Collaborate closely with experts across disciplines to work across the entire stack to improve the reliability and performance of the system.