Applied AI Scientist – Autograder for Image Quality

Socket.dev; Cupertino (CA); On-site; USD 150,000 - 210,000; Full time

Job summary

Apple seeks an AI Scientist to join a centralized evaluation org, building autograders for image generation features. You will craft scoring rubrics, train graders and validate outputs at scale, impacting AI experiences for hundreds of millions of users.

You will collaborate with MLEs, data teams and engineers to refine prompts, measure quality, and deploy autograder systems across workflows.

Qualifications

• Master's or PhD in CS/ML/AI or related field.
• Deep understanding of visual-language models.
• Proficiency in Python with production-ready code.
• Strong communication and stakeholder collaboration skills.

Responsibilities

• Focus on autograder research, training and adoption.
• Collaborate with AI feature teams and annotation teams to refine requirements.
• Develop, evaluate and iterate on grading prompts to align autograder with rubrics.
• Apply advanced techniques like fine-tuning to close grading gaps.
• Design analyses to measure and explain autograder quality.
• Collaborate with MLEs and teams on autograder deployment and adoption.
• Develop scalable systems to speed up autograder training processes.

Skills

Python

Visual-language models

Communication

Evaluation design

Education

Master's or PhD in Computer Science, Machine Learning, AI or related

Job description

Apple seeks an AI Scientist to join a centralized evaluation org, building autograders for image generation features. You will craft scoring rubrics, train graders and validate outputs at scale, impacting AI experiences for hundreds of millions of users.

You will collaborate with MLEs, data teams and engineers to refine prompts, measure quality, and deploy autograder systems across workflows.