Key Skills

Software

Appen

Data Annotation Tech

Labelbox

Mindrift

Remotasks

Scale AI

Top Subject Matter

No subject matter listed

Top Data Types

Audio

Computer Code Programming

Text

Top Task Types

Audio Recording

Classification

Computer Programming Coding

Evaluation Rating

Prompt Response Writing SFT

Freelancer Overview

I have 3 years of hands-on experience in AI training and data annotation across multiple domains, including natural language processing, voice, and programming. At Outlier AI, I contributed to a variety of projects such as Arabic and English AI training, math data annotation, and Arabic voice AI training using real-time feedback (RTF). My work focused on ensuring high-quality, culturally accurate, and well-structured training data for AI models. I also worked at Alignerr as a coding AI trainer, where I specialized in Python, Go, HTML, C++, and Java. In addition to programming tasks, I contributed to Arabic AI training for both text and voice-based projects. My diverse experience in multilingual data annotation, technical training, and linguistic accuracy makes me a strong contributor to any AI development team.

Labeling Experience

Green Withards

Scale AITextQuestion AnsweringText Generation

In this project, we are required to create difficult math problems in different subjects that the model fails to answer. Then, we are required to solve these problems and evaluate and fix the model’s answers using step-by-step reasoning.

In this project, we are required to create difficult math problems in different subjects that the model fails to answer. Then, we are required to solve these problems and evaluate and fix the model’s answers using step-by-step reasoning.

2024 - 2024

Cypher RLHF

Scale AITextRLHFEvaluation Rating

The Cypher RLHF project is designed to enhance the performance and reliability of AI language models by evaluating how they respond to human-written prompts. The core idea is to identify weaknesses or limitations in AI-generated responses through targeted testing and human evaluation. Participants are asked to create prompts that intentionally lead the model to make mistakes or produce unsatisfactory answers. Then, they compare two different AI-generated responses, select the better one, and explain their choice. If the better response still contains errors, it is rewritten to fix all issues. This process supports Reinforcement Learning from Human Feedback (RLHF), helping AI models learn from human preferences and improve their natural language understanding, factual accuracy, and communication style over time.

The Cypher RLHF project is designed to enhance the performance and reliability of AI language models by evaluating how they respond to human-written prompts. The core idea is to identify weaknesses or limitations in AI-generated responses through targeted testing and human evaluation. Participants are asked to create prompts that intentionally lead the model to make mistakes or produce unsatisfactory answers. Then, they compare two different AI-generated responses, select the better one, and explain their choice. If the better response still contains errors, it is rewritten to fix all issues. This process supports Reinforcement Learning from Human Feedback (RLHF), helping AI models learn from human preferences and improve their natural language understanding, factual accuracy, and communication style over time.

2024 - 2024

Education

Work History