Mohammed B.

Key Skills

Software

Appen

Don't disclose

Top Subject Matter

LLM response evaluation and quality alignment

Web search relevance rating and content safety

AI evaluation annotations for text and search datasets

Top Data Types

Text

Image

Top Task Types

Classification

Freelancer Overview

AI Data Trainer & Evaluator (LLM multi-turn evaluation, red-teaming, and labeling). Brings 1+ years of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include Internal, Proprietary Tooling, and EWOQ. Education includes Bachelor of Commerce, Cairo University (2013). AI-training focus includes data types such as Text and labeling workflows including Evaluation, Rating, and Classification.

Labeling Experience

AI Model Evaluator & Software Data Rater (Turing)

Don't discloseText

Trained and evaluated advanced AI model responses through structured testing focused on clarity, accuracy, safety, and helpfulness. Reviewed and labeled complex datasets within remote evaluation workflows to optimize model behavior for real-world deployment. Applied consistent labeling criteria aligned to safety and usefulness requirements.• Rated responses for helpfulness, clarity, and correctness.• Assessed safety constraints and compliance of outputs.• Labeled complex dataset examples for model improvement.• Collaborated in remote workflows to meet evaluation standards.

Present

Data Rater & Annotation Specialist (CrowdGen/Appen)

AppenTextClassification

Performed data labeling and content categorization for large-scale AI benchmarking projects. Delivered consistent tagging and categorization to help train machine learning models for semantic understanding and search precision. Worked through defined workflows to ensure labeled outputs matched expected schemas and criteria.• Labeled and categorized dataset items for benchmarking tasks.• Produced semantic tags to support model training.• Ensured consistent labeling quality across workflows (e.g., Echo, Multimango).• Contributed to improved search relevance and semantic modeling.

Present

AI Community Evaluator & Rater

Text

Participated in AI evaluation programs by providing precise annotations for text, search queries, and digital media datasets. Managed high-volume task queues on designated platforms while meeting strict internal QA thresholds. Flagged algorithmic inconsistencies to support refinement of localized search and data processing models.• Annotated text and search-query related dataset items for evaluation.• Completed tasks in high-volume queues with QA compliance.• Investigated and reported inconsistencies affecting model behavior.• Supported iteration of localized search/data processing systems.

Present

Search Quality Rater & Relevance Specialist

Text

Analyzed and evaluated web search results, advertisements, and user queries to improve search relevance. Interpreted user intent across varied contexts to produce accurate ratings for relevance, utility, and content safety. Maintained high production speeds while adapting to frequent and detailed guideline updates.• Rated search and ad relevance based on user intent signals.• Assessed utility and safety of content in search outputs.• Followed rapidly changing, detailed platform guidelines.• Delivered consistent quality scores under throughput targets.

Present

AI Data Trainer & Evaluator (LLM multi-turn evaluation, red-teaming, and labeling)

Text

Evaluated, ranked, and human-verified multi-turn LLM completions against strict engineering guidelines. Performed adversarial red-teaming by crafting prompting to expose weaknesses, then rewrote model responses to fix factual, logical, or stylistic issues. Completed specialized annotation and labeling tasks with deep analytical checks for formatting and quality compliance.• Ranked and verified complex assistant responses for quality and correctness.• Conducted red-teaming/adversarial prompting to identify failure cases.• Rewrote responses to improve accuracy, coherence, and guideline adherence.• Ensured outputs met formatting constraints and quality benchmarks.

Present

Education

Work History