Marwan Sherif

Key Skills

Software

Mercor

Other

Appen

Clickworker

CrowdSource

Data Annotation Tech

HiveMind

Labelbox

Micro1

Mindrift

OneForma

Remotasks

Scale AI

SuperAnnotate

Toloka

Telus

Top Subject Matter

Artificial Intelligence, Safety & LLM Evaluation

Language & Linguistics (English–Arabic Translation and Transcription)

Healthcare / Medical Sciences (Dentistry & Clinical Terminology)

Top Data Types

Text

Medical Dicom

Top Task Types

Red Teaming

Segmentation

Classification

Bounding Box

Object Detection

Text Generation

Question Answering

Text Summarization

RLHF

Fine Tuning

Transcription

Evaluation Rating

Data Collection

Prompt Response Writing SFT

Freelancer Overview

AI Red Teaming Specialist – Mercor. Brings 8+ years of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include Mercor and Other. Education includes Bachelor of Dental Surgery, Misr International University (2020) and Master of Endodontics, Misr International University (2022). AI-training focus includes data types such as Text and labeling workflows including Red Teaming.

Labeling Experience

AI Red Teaming Specialist – Mercor

MercorTextRed Teaming

As an AI Red Teaming Specialist at Mercor, I conducted adversarial testing of large language models to identify safety, alignment, and policy compliance vulnerabilities. I designed structured red-team attack scenarios and evaluated model responses for robustness, hallucination rates, and bias exposure. I documented reproducible failure cases, providing clear annotations to support model improvement and safety advancements. • Designed and executed red-teaming and adversarial prompt testing. • Analyzed model outputs for robustness, refusal, bias, and hallucination. • Annotated vulnerabilities and provided feedback to enhance AI safety. • Utilized internal/proprietary red-teaming and evaluation frameworks.

As an AI Red Teaming Specialist at Mercor, I conducted adversarial testing of large language models to identify safety, alignment, and policy compliance vulnerabilities. I designed structured red-team attack scenarios and evaluated model responses for robustness, hallucination rates, and bias exposure. I documented reproducible failure cases, providing clear annotations to support model improvement and safety advancements. • Designed and executed red-teaming and adversarial prompt testing. • Analyzed model outputs for robustness, refusal, bias, and hallucination. • Annotated vulnerabilities and provided feedback to enhance AI safety. • Utilized internal/proprietary red-teaming and evaluation frameworks.

2026 - Present

LLM Tester / AI Red Teaming Contributor – Turing

OtherTextRed Teaming

As an LLM Tester and AI Red Teaming Contributor at Turing, I crafted adversarial prompts to induce constraint violations, unsafe compliance, and hallucinations. I performed structured multi-turn attacks and evaluated outputs against strict safety and accuracy guidelines. I identified and documented recurrent failure patterns to improve model performance and safety. • Created and applied adversarial prompt attacks. • Systematically evaluated model responses for safety and compliance. • Provided detailed annotation and documentation on model vulnerabilities. • Collaborated using internal evaluation tools and frameworks.

As an LLM Tester and AI Red Teaming Contributor at Turing, I crafted adversarial prompts to induce constraint violations, unsafe compliance, and hallucinations. I performed structured multi-turn attacks and evaluated outputs against strict safety and accuracy guidelines. I identified and documented recurrent failure patterns to improve model performance and safety. • Created and applied adversarial prompt attacks. • Systematically evaluated model responses for safety and compliance. • Provided detailed annotation and documentation on model vulnerabilities. • Collaborated using internal evaluation tools and frameworks.

2024 - Present

Applied AI Language & Adversarial Evaluation – Independent

OtherTextRed Teaming

As an independent practitioner, I applied AI language evaluation and adversarial testing to assess LLMs and NLP systems. I tested for linguistic ambiguity, context sensitivity, bias, and vulnerability to unsafe outputs. I annotated and documented reproducible edge cases to support model safety enhancements. • Developed prompts targeting linguistic failure modes and edge cases. • Annotated vulnerabilities and unsafe outputs for NLP systems. • Supported model improvement with structured documentation of failures. • Used internal and open-source tools for prompt engineering and evaluation.

As an independent practitioner, I applied AI language evaluation and adversarial testing to assess LLMs and NLP systems. I tested for linguistic ambiguity, context sensitivity, bias, and vulnerability to unsafe outputs. I annotated and documented reproducible edge cases to support model safety enhancements. • Developed prompts targeting linguistic failure modes and edge cases. • Annotated vulnerabilities and unsafe outputs for NLP systems. • Supported model improvement with structured documentation of failures. • Used internal and open-source tools for prompt engineering and evaluation.

2020 - Present

Education

Work History