About Turing
Turing is an AI-powered tech services company with a mission to accelerate AGI advancement and deployment by bridging the gap between global talent from 100+ countries and the world's best foundational LLM companies, by helping them improve performance through model evaluation, fine-tuning, feedback, factuality and handling diverse data.
Role Overview
We are seeking detail-oriented QA Reviewers to check and approve image edits used to train next-generation AI image editing models. Each task contains an original image, a written editing instruction (prompt), an edit category (L2) and the edited image produced by an annotator. You verify that the edit follows the instruction exactly, looks photorealistic and leaves everything else untouched, then approve it, fix it or send it back with clear feedback.
This role suits senior photo retouchers, art directors, creative leads and image QA specialists who combine a sharp eye for visual defects with strong written English. It is a remote, project-based role, with training and guidelines provided.
What does day-to-day look like
- Review edited images against the original image, the prompt and its L2 category, and confirm the edit does exactly what was asked, nothing more and nothing less
- Check realism and technical quality: lighting, shadows, reflections, perspective, edges, color consistency, and no halos, seams, warping or artifacts
- Confirm that areas the instruction does not mention are unchanged, including identity, pose, background and resolution
- Score each task against the project quality rubric and mark it as approved, minor fix or rework
- Write constructive, specific feedback and detailed annotations so annotators know exactly what to correct
- Flag prompts that are unclear, impossible or inconsistent with their L2 category, and escalate recurring issues to the project lead
- Spot error patterns across annotators and contribute to guideline updates and calibration sessions
- Meet daily review throughput and accuracy targets while keeping scoring consistent with other reviewers