3 months ago
Responsibilities
- Contribute to the technical direction of the Perception Architecture team and broader perception organization.
- Design and develop core machine learning components and architectures for scalable, reliable perception in complex long-tail driving scenarios.
- Prototype key architectural decisions, including approaches to long-tail challenges and long-range detection.
- Improve model frameworks, architectures, and data pipelines to reduce development bottlenecks and accelerate model iteration.
- Lead challenging technical topics and build consensus across perception teams and stakeholders.
Requirements
- BS with 4+ years of experience, MS with 2+ years of experience, or PhD with 1+ years of experience in a perception-related field such as Robotics, Computer Vision, or Machine Learning.
- Extensive experience architecting, training, and evaluating deep learning models for challenging problem spaces.
- A track record of driving applied research or engineering projects from conception and experimentation through successful productization.
- Strong software engineering, machine learning algorithm design, and ML data and metrics pipeline development experience.
- Fluency in Python and hands-on experience with C++.
- Experience delivering detection, segmentation, tracking, or other perception solutions for real-time robotic applications; sensor fusion experience is highly desirable.
- Candidates may be subject to U.S. person, citizenship, residence, export-control, and national-security eligibility restrictions.
Categories
ML EngineeringRobotics
About Stack AV
Stack AV builds autonomous driving software and integrated systems for commercial trucking, combining AI perception, planning, robotics, and cloud-based fleet operations. The company works with freight carriers and logistics operators to pilot and deploy autonomy-enabled trucking solutions and supporting services. Founded in 2023 and headquartered in Pittsburgh, PA, it is a privately held firm focused on long-haul and hub-to-hub routes.