Mehran Amin

Key Skills

Software

Appen

Other

Top Subject Matter

No subject matter listed

Top Data Types

Document

Text

Top Label Types

Evaluation Rating

Prompt Response Writing SFT

Transcription

Translation Localization

Freelancer Overview

I am an electronics and communication engineering student with a strong foundation in machine learning, data analysis, and programming using Python and C. My hands-on experience includes building predictive models, performing data preprocessing, feature selection, and model evaluation on real-world datasets. I am proficient with data analysis libraries such as NumPy and Pandas, and have applied these skills in projects like movie recommendation systems and student performance prediction. My technical background also covers MATLAB-based signal processing and practical circuit design, giving me a unique perspective on integrating data-driven approaches with engineering tasks. I am detail-oriented, analytical, and eager to contribute to data labeling and AI training data initiatives, ensuring high-quality, accurate datasets for machine learning applications.

Labeling Experience

AI Response Evaluator

AppenTextTranslation Localization

Worked on language quality and translation evaluation projects focused on assessing AI-generated and human-translated content. Responsibilities included reviewing source and target texts to ensure semantic accuracy, linguistic fidelity, and tone preservation across languages. Evaluated translations against strict quality guidelines, identifying issues such as mistranslation, omission, added meaning, register mismatch, and contextual errors. Provided structured feedback and quality ratings to support the improvement of multilingual AI systems, with a strong emphasis on consistency, instruction adherence, and cultural appropriateness. Collaborated within a guideline-driven evaluation framework where precision, objectivity, and attention to detail were critical to maintaining high standards of language performance.

Worked on language quality and translation evaluation projects focused on assessing AI-generated and human-translated content. Responsibilities included reviewing source and target texts to ensure semantic accuracy, linguistic fidelity, and tone preservation across languages. Evaluated translations against strict quality guidelines, identifying issues such as mistranslation, omission, added meaning, register mismatch, and contextual errors. Provided structured feedback and quality ratings to support the improvement of multilingual AI systems, with a strong emphasis on consistency, instruction adherence, and cultural appropriateness. Collaborated within a guideline-driven evaluation framework where precision, objectivity, and attention to detail were critical to maintaining high standards of language performance.

2025

Response Evaluator

OtherTextEvaluation Rating

Evaluated AI-generated responses in the Urdu language for accuracy, relevance, fluency, and adherence to task-specific guidelines. Responsibilities included comparing multiple model outputs, selecting the most accurate response, assigning quality scores using defined rubrics, and providing clear justifications for evaluation decisions. Focused on identifying issues such as mistranslation, incorrect or added meaning, grammatical and stylistic errors, tone mismatch, and failure to follow instructions. Ensured fidelity to the original prompt without assumptions or hallucinated context, maintaining consistency and objectivity across evaluations. Contributed to improving the linguistic quality and reliability of AI systems through detailed, guideline-driven feedback.

Evaluated AI-generated responses in the Urdu language for accuracy, relevance, fluency, and adherence to task-specific guidelines. Responsibilities included comparing multiple model outputs, selecting the most accurate response, assigning quality scores using defined rubrics, and providing clear justifications for evaluation decisions. Focused on identifying issues such as mistranslation, incorrect or added meaning, grammatical and stylistic errors, tone mismatch, and failure to follow instructions. Ensured fidelity to the original prompt without assumptions or hallucinated context, maintaining consistency and objectivity across evaluations. Contributed to improving the linguistic quality and reliability of AI systems through detailed, guideline-driven feedback.

2024 - 2025

Education

Work History