Muhammad Ghazal

Key Skills

Software

Appen

Clickworker

Data Annotation Tech

Mercor

Other

Top Subject Matter

No subject matter listed

Top Data Types

Audio

Image

Text

Top Label Types

Evaluation Rating

Data Collection

Prompt Response Writing SFT

Freelancer Overview

I am a native Arabic speaker and bilingual specialist with hands-on experience in data annotation, linguistic evaluation, and AI training data workflows. My background includes working on projects such as Nightshade and Argon, where I developed rubrics, performed conversational testing, and evaluated AI-generated responses for correctness and cultural relevance in both Modern Standard Arabic and Egyptian dialects. I am skilled in using annotation tools, conducting manual data labeling, and ensuring high-quality, contextually accurate outputs while strictly following project guidelines. My strong attention to detail, fast learning ability, and clear communication in both Arabic and English enable me to adapt quickly to new tasks, including audio and video verification, and contribute effectively to remote QA teams.

Labeling Experience

English Voice Recording Podcasts Project

OtherAudioAudio Recording

Supported podcast production by recording English narration from written scripts and ensuring delivery-ready audio assets. Collaborated asynchronously with the production team, applied feedback to retakes, and followed a structured workflow to maintain consistent audio quality, accurate pronunciation, and timely delivery across episodes.

Supported podcast production by recording English narration from written scripts and ensuring delivery-ready audio assets. Collaborated asynchronously with the production team, applied feedback to retakes, and followed a structured workflow to maintain consistent audio quality, accurate pronunciation, and timely delivery across episodes.

2025 - 2025

The Plinia Project

Data Annotation TechImageEvaluation RatingData Collection

Plinia is an AI dataset development project centered on generating AI-driven image prompts and golden responses in Egyptian Arabic to enrich training data with culturally aligned, high-quality examples.  In this project, I owned the prompt and golden-response creation pipeline, deriving reliable interpretations from visual analysis and turning them into natural-language outputs that reflect Egyptian cultural context and everyday usage. My work prioritized cultural relevance, clarity, and consistency, ensuring each prompt-response pair met quality expectations and supported stronger user engagement through realistic, context-aware content.

Plinia is an AI dataset development project centered on generating AI-driven image prompts and golden responses in Egyptian Arabic to enrich training data with culturally aligned, high-quality examples.  In this project, I owned the prompt and golden-response creation pipeline, deriving reliable interpretations from visual analysis and turning them into natural-language outputs that reflect Egyptian cultural context and everyday usage. My work prioritized cultural relevance, clarity, and consistency, ensuring each prompt-response pair met quality expectations and supported stronger user engagement through realistic, context-aware content.

2025 - 2025

The Argon Chatbot Project

Data Annotation TechTextEvaluation Rating

Argon is a Modern Standard Arabic (MSA) chatbot evaluation project designed to ensure consistently high-quality, contextually appropriate conversational performance.  I led conversational testing and response editing and developed an evaluation rubric with explicit benchmarks to assess interaction quality and linguistic accuracy across diverse real-world scenarios. My work focused on refining chatbot behavior through structured dialogue testing, identifying conversational breakdowns (e.g., context drift, incomplete answers, phrasing issues), and enhancing outputs to better align with MSA conventions and user intent, resulting in more reliable natural interactions.

Argon is a Modern Standard Arabic (MSA) chatbot evaluation project designed to ensure consistently high-quality, contextually appropriate conversational performance.  I led conversational testing and response editing and developed an evaluation rubric with explicit benchmarks to assess interaction quality and linguistic accuracy across diverse real-world scenarios. My work focused on refining chatbot behavior through structured dialogue testing, identifying conversational breakdowns (e.g., context drift, incomplete answers, phrasing issues), and enhancing outputs to better align with MSA conventions and user intent, resulting in more reliable natural interactions.

2025 - 2025

Nightshade Project

Data Annotation TechTextEvaluation Rating

Nightshade is an AI/NLP evaluation project designed to elevate the quality of AI-generated responses for Egyptian Arabic users. In this project, I led the evaluation workflow and developed the scoring rubric, establishing a structured quality standard to assess AI outputs for correctness, factuality, and completeness. My work emphasized high-precision judgment under strict guidelines, ensuring consistent evaluation decisions and reliable quality signals for model improvement. I applied a methodical approach, similar to engineering QA practices, to identify response gaps, categorize issues, and strengthen evaluation criteria for Arabic-Egyptian linguistic and contextual accuracy.

Nightshade is an AI/NLP evaluation project designed to elevate the quality of AI-generated responses for Egyptian Arabic users. In this project, I led the evaluation workflow and developed the scoring rubric, establishing a structured quality standard to assess AI outputs for correctness, factuality, and completeness. My work emphasized high-precision judgment under strict guidelines, ensuring consistent evaluation decisions and reliable quality signals for model improvement. I applied a methodical approach, similar to engineering QA practices, to identify response gaps, categorize issues, and strengthen evaluation criteria for Arabic-Egyptian linguistic and contextual accuracy.

2025 - 2025

Education

Work History