Ksenia Sterkhova

Key Skills

Software

Remotasks

Scale AI

Top Subject Matter

No subject matter listed

Top Data Types

Text

Top Task Types

Prompt Response Writing SFT

Question Answering

Text Generation

Text Summarization

Translation Localization

Freelancer Overview

I have experience working on AI training data projects where I helped improve Large Language Models through data labeling, evaluation, and quality checks. As a Queue Manager at Outlier/Scale AI, I supported contributors, kept them engaged, solved issues, and made sure tasks were completed with good quality. I also worked with teams and communities, helping communication between contributors and project managers. In addition to management, I took part in different AI projects myself. This includes writing rubrics to evaluate prompts and model answers, working on localization and cultural adaptation, checking language quality, reviewing safety and harmful content, and doing translation tasks. With this mix of management and hands-on project work, I can bring reliable skills to both organizing teams and creating high-quality AI training data.

Labeling Experience

Harmful Content & Safety Review

Scale AITextTranslation LocalizationEvaluation Rating

Reviewed AI-generated text for harmful, unsafe, or sensitive content, labeling it according to safety and policy guidelines. Evaluated outputs in multiple languages, providing translations and cultural adaptations to ensure accurate and context-appropriate responses. Combined content moderation with linguistic review to improve AI safety and multilingual understanding.

Reviewed AI-generated text for harmful, unsafe, or sensitive content, labeling it according to safety and policy guidelines. Evaluated outputs in multiple languages, providing translations and cultural adaptations to ensure accurate and context-appropriate responses. Combined content moderation with linguistic review to improve AI safety and multilingual understanding.

2025 - 2025

Rubric Writing & Evaluation

Scale AITextRLHFEvaluation Rating

Created and applied rubrics to evaluate prompts and AI model responses. Focused on clarity, correctness, and completeness of outputs. Helped improve model alignment by writing sample prompts and responses, and rating AI performance according to defined criteria.

Created and applied rubrics to evaluate prompts and AI model responses. Focused on clarity, correctness, and completeness of outputs. Helped improve model alignment by writing sample prompts and responses, and rating AI performance according to defined criteria.

2025 - 2025

Education

Work History