Staff Computer Vision Engineer, Perception

Metropolis•Seattle, WA

2d•$220,000 - $250,000•Onsite

About The Position

Metropolis is looking for a Staff Computer Vision Engineer to set technical direction and to lead the design and development of computer vision (CV) algorithms and systems that power Metropolisʼs “AI for the Real World”. We are developing a broad suite of CV-enabled systems that span both edge and cloud environments, and enable checkout-free experiences in a variety of applications (e.g. parking, fueling, drive-thrus). As a senior technical leader within the Advanced Technologies Group (ATG), you will drive the architecture, optimization, integration, and validation of advanced perception models for deployment on edge and cloud platforms. This will include solving complex challenges in object detection, tracking, and recognition, and may span video analytics and multi-modal systems. You will work closely with cross-functional teams throughout the entire perception system lifecycle, from development to rigorous testing, while driving innovations in computer vision and machine learning algorithms that directly enhance the accuracy, reliability, and performance of applications core to our company. Excellence in this position involves leading innovation in computer vision, delivering impactful results, and fostering growth through technical leadership and mentorship within the ML team. You must be able to thrive in a entrepreneurial environment, working collaboratively in a fast-paced environment that often has multiple stakeholders. You wonʼt be afraid to break new technological ground at Metropolis, and are more than willing to roll up your sleeves, dig in, and get the job done.

Requirements

M.S. (with 8+ years of industry experience) or Ph.D. (with 5+ years of industry experience) in Computer Science, Electrical Engineering, Robotics, or related field with a focus on computer vision or machine learning
Consistent track record of developing and deploying perception algorithms for real-world applications (e.g. autonomous driving) to deliver business impact
Strong proficiency in modern deep learning frameworks (e.g., PyTorch, TensorFlow, Qualcomm Chip) and classical CV libraries (e.g., OpenCV)
Deep expertise in two or more of: (2D/3D) object detection, multi-object tracking, segmentation, depth estimation, localization, or sensor fusion
Solid experience with deployment of ML models on edge devices and Cloud Platform (e.g., GPU/embedded)
Excellent programming skills (5+ years or more) in Python and/or C++
Strong experience (3+ years or more) with ML model training, Data Annotation and Model Evaluation
Strong understanding of the agile development process and CI/CD pipelines and tools (e.g., Github Actions, Jenkins)

Nice To Haves

Prior experience leading computer vision teams and owning production deployments highly desirable
Publications (CVPR, NeurIPS, ICML, etc) or patents in computer vision or applied machine learning preferred

Responsibilities

Architect and develop advanced computer vision algorithms for tasks such as object detection, tracking, semantic segmentation, and scene understanding
Take initiative to identify gaps, bottlenecks, and opportunities in system design and performance —then drive solutions from concept through deployment
Define and implement key metrics and KPIs to rigorously evaluate system accuracy, robustness, and scalability across diverse real-world conditions
Drive system-level debugging and root cause analysis for perception failures in production environments
Collaborate across hardware, ML, and software engineering teams to improve data pipelines, model deployment frameworks, and sensor configurations.
Mentor and lead by example, promoting best practices in scalable architecture, model evaluation, and engineering excellence within the computer vision team
Research, prototype, and integrate new machine learning or deep learning algorithms applicable to perception problems
Participate in cross-team initiatives to improve system reliability, labeling quality, simulation- based validation, optimizing perception models and production-readiness of perception models
Stay ahead of the curve by researching, prototyping, and integrating emerging deep learning techniques and tools into our perception stack