Tarek Elgebily

Key Skills

Software

Appen

Labelbox

OneForma

Scale AI

Other

Top Subject Matter

No subject matter listed

Top Data Types

Audio

Text

Top Task Types

Evaluation Rating

Prompt Response Writing SFT

RLHF

Text Generation

Text Summarization

Freelancer Overview

With over a year of experience in data labeling and AI training, I have specialized in optimizing large language models (LLMs) with a focus on Egyptian, Saudi, and Emirati Arabic dialects. My expertise includes extensive work in Reinforcement Learning with Human Feedback (RLHF), where I have trained models to improve accuracy, contextual understanding, and user alignment through careful evaluation and iterative feedback processes. As a Quality Manager in transcription projects, I have led teams to refine workflows and implement high-quality standards in speech recognition, consistently achieving an average feedback rating of 4/5 in LLM module evaluations. What sets me apart is my ability to bridge technical expertise with domain-specific knowledge, including a pharmaceutical background that enables precise applications in healthcare-related AI. I excel in cultural and linguistic localization and have proven success in delivering AI solutions tailored to diverse audiences. My hands-on experience in RLHF techniques, coupled with leadership in data labeling workflows, makes me a strong contributor to projects requiring strategic and detail-oriented AI training approaches.

Labeling Experience

AI Evaluator and Quality Specialist

AppenTextText GenerationText Summarization

Contributed to a comprehensive project aimed at fine-tuning large language models (LLMs) across multiple linguistic and cultural contexts. Responsibilities included creating and evaluating prompts, providing detailed feedback on AI-generated responses, and conducting preference rankings to optimize model performance. Focused on enhancing model capabilities in adhering to task-specific instructions, ensuring accuracy, coherence, and contextual relevance, while adhering to strict confidentiality and project guidelines.

Contributed to a comprehensive project aimed at fine-tuning large language models (LLMs) across multiple linguistic and cultural contexts. Responsibilities included creating and evaluating prompts, providing detailed feedback on AI-generated responses, and conducting preference rankings to optimize model performance. Focused on enhancing model capabilities in adhering to task-specific instructions, ensuring accuracy, coherence, and contextual relevance, while adhering to strict confidentiality and project guidelines.

2024

AI Model Trainer – RLHF Specialist

Scale AITextText GenerationRLHF

Collaborated on a high-impact project focused on enhancing the performance of large language models through Reinforcement Learning with Human Feedback (RLHF). Tasks included prompt creation, response evaluation, and preference ranking to improve model alignment with user needs. Played a key role in ensuring that AI responses adhered to linguistic accuracy, cultural relevance, and task-specific requirements while maintaining confidentiality and adhering to strict project guidelines.

Collaborated on a high-impact project focused on enhancing the performance of large language models through Reinforcement Learning with Human Feedback (RLHF). Tasks included prompt creation, response evaluation, and preference ranking to improve model alignment with user needs. Played a key role in ensuring that AI responses adhered to linguistic accuracy, cultural relevance, and task-specific requirements while maintaining confidentiality and adhering to strict project guidelines.

2024

Scripted Monologue Audio Validation - Sportscast

OtherAudioClassificationEmotion Recognition

Validating audio recordings from native English (GB) speakers to ensure they match the given prompts and have the right expressiveness for the Sportscast domain. Tasks include listening to the audio clips, evaluating attributes such as emotion, disfluency, and background noise, and making any necessary corrections to the text.

Validating audio recordings from native English (GB) speakers to ensure they match the given prompts and have the right expressiveness for the Sportscast domain. Tasks include listening to the audio clips, evaluating attributes such as emotion, disfluency, and background noise, and making any necessary corrections to the text.

2024 - 2024

Audio Transcription Specialist – Multilingual Speech Recognition Project

AppenAudioClassificationTranslation Localization

Participated in a specialized project to enhance automatic speech recognition models by accurately transcribing and tagging audio in Egyptian Arabic. Responsibilities included correcting pre-filled transcriptions, applying contextual tagging for speech and non-speech sounds, and maintaining strict adherence to guidelines for timestamping and speaker diarization. Delivered high-quality results using advanced transcription tools, ensuring precision and consistency in alignment with project standards and confidentiality requirements.

Participated in a specialized project to enhance automatic speech recognition models by accurately transcribing and tagging audio in Egyptian Arabic. Responsibilities included correcting pre-filled transcriptions, applying contextual tagging for speech and non-speech sounds, and maintaining strict adherence to guidelines for timestamping and speaker diarization. Delivered high-quality results using advanced transcription tools, ensuring precision and consistency in alignment with project standards and confidentiality requirements.

2024 - 2024

Education

Work History