Taras Tsohla

Key Skills

Software

Scale AI

Top Subject Matter

No subject matter listed

Top Data Types

Text

Top Label Types

Evaluation Rating

Prompt Response Writing SFT

Question Answering

RLHF

Freelancer Overview

I am an experienced data annotator and AI trainer with a strong background in multilingual text analysis, data quality assurance, and human-in-the-loop AI training. I have hands-on expertise in annotating, classifying, validating, and improving AI-generated content across domains such as linguistics, social sciences, and policy analysis. My work includes large-scale text classification, intent detection, relevance ranking, and detailed evaluation of AI outputs for accuracy, bias, and guideline compliance. Fluent in English, Slovak, Ukrainian, and Russian, I excel at following complex annotation guidelines, ensuring consistency, and delivering high-quality, structured data for AI model training. I am skilled in both qualitative and quantitative analysis, and I routinely provide feedback to enhance model reasoning and performance while maintaining the highest standards of data integrity and confidentiality.

Labeling Experience

Evaluations Cypher evals

Scale AITextRLHFEvaluation Rating

Cypher Evals — an AI response evaluation project. Participants analyze two AI-generated outputs for accuracy, safety, and localization, then rank them and provide written justifications, helping improve the model’s performance and reliability.

Cypher Evals — an AI response evaluation project. Participants analyze two AI-generated outputs for accuracy, safety, and localization, then rank them and provide written justifications, helping improve the model’s performance and reliability.

2025

Prompt creation - NDA name

Scale AITextPrompt Response Writing SFT

an AI training project focused on creating diverse prompt sets incorporating culturally specific topics relevant to particular languages and regions. Participants designed prompts reflecting local context, social norms, and linguistic nuances, enabling more accurate, culturally aware, and localized AI responses.

an AI training project focused on creating diverse prompt sets incorporating culturally specific topics relevant to particular languages and regions. Participants designed prompts reflecting local context, social norms, and linguistic nuances, enabling more accurate, culturally aware, and localized AI responses.

2025 - 2025

Education

Work History