Balkis Hasanah

Key Skills

Software

Don't disclose

iMerit

Other

Top Subject Matter

Search Engine Quality and AI Relevance Assessment

Large Language Model (LLM) Quality Assessment

Computer Vision and Image Annotation

Top Data Types

Text

Image

Document

Top Task Types

Classification

Evaluation Rating

Freelancer Overview

AI Quality Evaluator with hands-on experience in search relevance rating, data annotation, and LLM evaluation. Skilled in applying Page Quality, Needs Met, and E-E-A-T frameworks, conducting side-by-side comparisons, and identifying hallucinations to ensure accurate, safe, and high-quality AI outputs. Experienced in handling ambiguity, analyzing user intent, and delivering consistent, guideline-based evaluations. Detail-oriented and reliable, with strong proficiency in Indonesian and English, committed to improving AI model performance through high-quality training data.

Labeling Experience

Quality Rater — Search Evaluation Specialist

TextEvaluation Rating

As a Quality Rater at Welocalize, I evaluated search results for relevance and accuracy using established frameworks. I performed data annotation and content quality assessment to support AI and machine learning training. I provided structured, guideline-based evaluations and detailed feedback for improving AI algorithms and search engine performance. • Utilized Page Quality, Needs Met, and E-E-A-T assessment frameworks. • Conducted side-by-side comparison and user intent analysis for comprehensive evaluation. • Used internal/proprietary tools for strict guideline-based labeling and model improvement. • Regularly identified harmful or misleading content to enhance search quality and safety.

As a Quality Rater at Welocalize, I evaluated search results for relevance and accuracy using established frameworks. I performed data annotation and content quality assessment to support AI and machine learning training. I provided structured, guideline-based evaluations and detailed feedback for improving AI algorithms and search engine performance. • Utilized Page Quality, Needs Met, and E-E-A-T assessment frameworks. • Conducted side-by-side comparison and user intent analysis for comprehensive evaluation. • Used internal/proprietary tools for strict guideline-based labeling and model improvement. • Regularly identified harmful or misleading content to enhance search quality and safety.

2026 - Present

AI Trainer — Project Aether AI

Don T DiscloseTextPrompt Response Writing SFT

As an AI Trainer for Outlier's Project Aether AI, I contributed to large language model training through response evaluation. My responsibilities included comparing and ranking model outputs and identifying issues such as hallucinations and logical inconsistencies. I generated structured rationales to enhance model refinement and system performance. • Evaluated factual accuracy, content safety, and instruction compliance of LLM responses. • Used side-by-side (SxS) methodology and evidence-based justifications for output ranking. • Flagged harmful or logically flawed AI outputs for retraining. • Worked remotely using proprietary or undisclosed platforms.

As an AI Trainer for Outlier's Project Aether AI, I contributed to large language model training through response evaluation. My responsibilities included comparing and ranking model outputs and identifying issues such as hallucinations and logical inconsistencies. I generated structured rationales to enhance model refinement and system performance. • Evaluated factual accuracy, content safety, and instruction compliance of LLM responses. • Used side-by-side (SxS) methodology and evidence-based justifications for output ranking. • Flagged harmful or logically flawed AI outputs for retraining. • Worked remotely using proprietary or undisclosed platforms.

2026 - 2026

Content Moderator — Live Streaming Gaming

VideoClassification

As a Content Moderator at Gear Inc, I assessed both AI-generated and user-generated textual content in live streaming gaming environments. I identified hallucinations, ambiguities, and assessed the quality of conversational outputs for policy compliance. I performed side-by-side model comparisons and provided structured ranking justifications for AI improvement. • Conducted detailed content evaluation and ambiguity resolution for conversational AI outputs. • Used structured frameworks to ensure natural and policy-compliant response generation. • Maintained high data confidentiality and standardized review documentation. • Improved model accuracy by documenting reasoning and decision patterns.

As a Content Moderator at Gear Inc, I assessed both AI-generated and user-generated textual content in live streaming gaming environments. I identified hallucinations, ambiguities, and assessed the quality of conversational outputs for policy compliance. I performed side-by-side model comparisons and provided structured ranking justifications for AI improvement. • Conducted detailed content evaluation and ambiguity resolution for conversational AI outputs. • Used structured frameworks to ensure natural and policy-compliant response generation. • Maintained high data confidentiality and standardized review documentation. • Improved model accuracy by documenting reasoning and decision patterns.

2025 - 2025

Data Annotator — Project Image Annotation

ImeritImageClassification

As a Data Annotator for iMerit Scholars, I performed detailed image annotation for a computer vision project. Tasks involved labeling and classifying objects within images to meet strict quality standards. I collaborated remotely to achieve annotation targets and support AI model development. • Applied established annotation guidelines for image datasets. • Ensured consistency and accuracy in object classification and labeling. • Managed annotation workflows and met project deadlines independently. • Supported computer vision model training with high-quality image labels.

As a Data Annotator for iMerit Scholars, I performed detailed image annotation for a computer vision project. Tasks involved labeling and classifying objects within images to meet strict quality standards. I collaborated remotely to achieve annotation targets and support AI model development. • Applied established annotation guidelines for image datasets. • Ensured consistency and accuracy in object classification and labeling. • Managed annotation workflows and met project deadlines independently. • Supported computer vision model training with high-quality image labels.

2025 - 2025

Education

Work History