Overview
We are seeking
Arabic (Levantine) AI Evaluation Specialists to help assess and improve the performance of advanced AI systems. In this role, you'll contribute directly to the evaluation and enhancement of large language models (LLMs) by testing how they understand, generate, and respond to Arabic content.
You will craft realistic scenarios, analyze model outputs for quality and safety, and help ensure the technology delivers accurate, culturally appropriate, and reliable results. Your insights will play a key role in shaping smarter AI experiences.
Project Details
Language: Native fluency in Levantin Arabic
Location: Remote-Egypt
Project Duration: 3 months
Pay Rate: $10 USD/Hour
Schedule: 40 hours a week. 8 hours per day Mon-Fri
Start Date: February 2nd
What You Will Do
- Conduct side-by-side comparisons of AI responses and rate their quality on a 1–5 scale based on established guidelines.
- Design scenario-based and edge-case prompts to evaluate model behavior, including tricky, ambiguous, or incomplete information situations.
- Assess outputs for instruction adherence, factual accuracy, tone, safety, and overall usefulness.
- Develop clear evaluation rubrics and criteria to ensure consistent scoring across tasks.
- Create reliable reference materials (articles, transcripts, reports, etc.) to serve as the source of truth for testing.
- Write well-structured gold standard responses that demonstrate the most accurate and helpful answer.
- Identify potential issues such as hallucinations, inconsistencies, or cultural/contextual mismatches.
Qualifications
- Bachelor's degree or equivalent experience in Linguistics, Computational Linguistics, Communications, Technical Writing, or a related analytical field.
- B2 or superior level of English.
- Native fluency in Modern Standard Arabic in Levantine dialect.
- Strong understanding of the distinction between Fusha and Ammiyya
- Proven experience in a role involving AI data annotation, content quality review, search quality rating, or prompt engineering.
- Ability to work independently and manage workflows effectively in a remote environment.
Nice to Have
- Multilingual proficiency in one or more Arabic dialects.
- Strong attention to detail and critical thinking to identify hallucinations and bias
- Familiarity with data annotation platforms and model evaluation tools.
- Experience in prompt engineering, AI evaluation, linguistic QA, or translation is a plus
- Cultural familiarity with regional norms and high-context communication styles, particularly in the GCC region.
Note: Please do not use VPNs or IP-masking tools during the recruitment process — our security system requires accurate regional verification.
