Role OverviewWe are looking for a Junior/Middle Computer Vision Engineer to support high-volume execution across data preparation, model training, and evaluation for an AI team working with large-scale image and video datasets. You will curate and manage annotation workflows, run model training and evaluation jobs, maintain benchmarks, and collaborate with senior engineers on failure-case analysis.
What You Will Do
- Curate large-scale image and video datasets, manage labeling processes and workflows, and ensure the highest standards for dataset quality; - Run model training and evaluation jobs, ensuring experiments are executed smoothly and efficiently; - Document training results, maintain ongoing evaluation benchmarks, and track model performance over time; - Collaborate with senior engineers to analyze model failure cases and identify areas for data or algorithmic improvement;
Why It Might Be a Fit
We offer a clear growth path into applied modeling or MLOps for an early-career engineer eager to build hands-on AI experience. A strong, demonstrable willingness to learn and adapt in a fast-paced, mentorship-driven environment is a must.
Requirements
- Authorized to work for ANY employer in the US
- 1 to 3 years of experience in software engineering, data science, machine learning, or a related field
- Degree in Computer Science, Data Science, Engineering, Mathematics, or a related discipline (or equivalent practical experience)
- Foundational coding skills in Python
- Foundational understanding of machine learning concepts and workflows
- Basic knowledge of computer vision principles (e.g., image processing, object detection basics)
- Basic familiarity with cloud environments and compute resources
- Upper-intermediate English level
Benefits
- Professional growth: Mentorship, TechTalks, and personalized growth roadmaps
- Competitive compensation: USD-based pay with education, fitness, and team activity budgets
- Exciting projects: Modern solutions with Fortune 500 and top product companies
- Flextime: Flexible schedule with remote and office options
]]>