Fatma Kasem

Key Skills

Software

Scale AI

Mercor

Top Subject Matter

Arabic LLM Evaluation & Prompt Engineering

Bilingual Arabic-English Video/Image AI Training

Generalist – Arabic & English LLM Evaluation

Top Data Types

Text

Image

Top Task Types

Bounding Box

Segmentation

Classification

Entity (NER) Classification

Object Detection

Text Generation

Question Answering

Text Summarization

RLHF

Fine-tuning

Red Teaming

Transcription

Evaluation/Rating

Prompt + Response Writing (SFT)

Function Calling

Polygon

Computer Programming/Coding

Freelancer Overview

AI Prompt Engineer & Linguistic Quality Specialist. Brings 3+ years of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include Internal, Proprietary Tooling, and Scale AI. Education includes Bachelor of Science, Menoufia University (2022). AI-training focus includes data types such as Text and Image and labeling workflows including Evaluation and Rating.

Labeling Experience

Generalist – English & Arabic (AI LLM Evaluation)

MercorText

Assessed and improved general conversational and written Arabic-English LLM outputs for fluency, tone, and cultural nuance. Delivered detailed bilingual prompt–response pairs and provided error feedback to support frontier LLM training for AI labs. Participated in live prompt evaluation workflows, directly impacting model performance and reliability. • Evaluated both written and spoken AI output for linguistic quality. • Addressed real user queries as part of model refinement. • Raised quality standards through context-aware feedback. • Facilitated bilingual model development for improved responses.

Assessed and improved general conversational and written Arabic-English LLM outputs for fluency, tone, and cultural nuance. Delivered detailed bilingual prompt–response pairs and provided error feedback to support frontier LLM training for AI labs. Participated in live prompt evaluation workflows, directly impacting model performance and reliability. • Evaluated both written and spoken AI output for linguistic quality. • Addressed real user queries as part of model refinement. • Raised quality standards through context-aware feedback. • Facilitated bilingual model development for improved responses.

2025 - Present

AI Training Contributor – Aether Project (Bilingual Arabic & English)

Scale AIImage

Labeled and evaluated bilingual Arabic–English AI training data for both video and image content as part of the Aether project. Assessed model output for accuracy, cultural appropriateness, and relevance, providing structured feedback for continuous improvement. Carried out attention-to-detail visual comparison and scripted dialogues for multimodal AI model training. • Evaluated and ranked AI-generated text and visual content. • Provided clarity and correctness feedback to enhance model precision. • Compared images to identify mismatches and errors. • Supported training with customized prompt–response pairs.

Labeled and evaluated bilingual Arabic–English AI training data for both video and image content as part of the Aether project. Assessed model output for accuracy, cultural appropriateness, and relevance, providing structured feedback for continuous improvement. Carried out attention-to-detail visual comparison and scripted dialogues for multimodal AI model training. • Evaluated and ranked AI-generated text and visual content. • Provided clarity and correctness feedback to enhance model precision. • Compared images to identify mismatches and errors. • Supported training with customized prompt–response pairs.

2025 - Present

AI Prompt Engineer & Linguistic Quality Specialist

Text

Evaluated large language model (LLM) outputs for Arabic linguistic quality, focusing on script correctness, grammar, and hallucination rate. Developed and applied structured error taxonomy, adversarial test suites, and standardized prompt frameworks to assess and enhance model reliability. Collaborated cross-functionally to document failure modes and improve prompt engineering strategies for culturally relevant and accurate AI content. • Consistently reviewed AI-generated text for clarity and regional appropriateness. • Designed linguistic best practices and quality control documentation. • Tested LLM responses against diverse Arabic dialect inputs. • Iterated on A/B tested prompt structures for output consistency.

Evaluated large language model (LLM) outputs for Arabic linguistic quality, focusing on script correctness, grammar, and hallucination rate. Developed and applied structured error taxonomy, adversarial test suites, and standardized prompt frameworks to assess and enhance model reliability. Collaborated cross-functionally to document failure modes and improve prompt engineering strategies for culturally relevant and accurate AI content. • Consistently reviewed AI-generated text for clarity and regional appropriateness. • Designed linguistic best practices and quality control documentation. • Tested LLM responses against diverse Arabic dialect inputs. • Iterated on A/B tested prompt structures for output consistency.

2025 - Present

Generative AI Intern

Text

Applied prompt engineering methods to Arabic and English generative text tasks, focusing on correctness, meaning preservation, and context. Designed prompts for content generation, summarization, and reasoning, systematically reviewing outputs for coherence and grammatical issues. Built and tested AI-generated text prototypes, experimenting with prompt chaining and quality improvement strategies. • Designed prompts for generative tasks targeting accuracy. • Identified and documented output errors in AI-generated content. • Developed prototype integrations with leading LLM APIs. • Analyzed outputs for context alignment and linguistic quality.

Applied prompt engineering methods to Arabic and English generative text tasks, focusing on correctness, meaning preservation, and context. Designed prompts for content generation, summarization, and reasoning, systematically reviewing outputs for coherence and grammatical issues. Built and tested AI-generated text prototypes, experimenting with prompt chaining and quality improvement strategies. • Designed prompts for generative tasks targeting accuracy. • Identified and documented output errors in AI-generated content. • Developed prototype integrations with leading LLM APIs. • Analyzed outputs for context alignment and linguistic quality.

2025 - 2025

Education

Work History