Hidnaz Abdulkhader

Key Skills

Software

OneForma

Scale AI

Top Subject Matter

No subject matter listed

Top Data Types

Audio

Image

Text

Video

Top Label Types

Segmentation

Evaluation Rating

Bounding Box

Point Key Point

Entity Ner Classification

Question Answering

Text Generation

RLHF

Prompt Response Writing SFT

Freelancer Overview

I am an AI Trainer with hands-on experience in evaluating and annotating data for video and image generation models, specializing in computer vision tasks such as object detection, product detection, and image editing assessments. My work involves providing structured human feedback, conducting head-to-head model comparisons, and ensuring high-quality labeling standards with a strong attention to visual detail and adherence to task-specific guidelines. I am skilled in using generative AI tools and have a proven ability to analyze and optimize data-driven processes, drawing on my background in e-commerce analytics and technical project management. My strengths include precise annotation, consistency in labeling, and a commitment to supporting the development of robust AI systems.

Labeling Experience

Generalist

Scale AITextQuestion AnsweringText Generation

Project Scope: LLM Evaluation, Rubric Design & Comparative Response Analysis As part of a multimodal AI development initiative, contributed to Large Language Model (LLM) alignment and performance enhancement through structured evaluation and human feedback workflows. Responsibilities included generating high-quality prompts and benchmark texts, designing detailed evaluation rubrics, and conducting head-to-head comparative assessments of model-generated responses. Performed qualitative and quantitative analysis of outputs based on coherence, factual accuracy, reasoning depth, instruction adherence, and safety compliance. Executed structured question-answering tasks to test domain adaptability and contextual understanding, while documenting edge cases and failure patterns to inform iterative fine-tuning and reinforcement learning cycles.

Project Scope: LLM Evaluation, Rubric Design & Comparative Response Analysis As part of a multimodal AI development initiative, contributed to Large Language Model (LLM) alignment and performance enhancement through structured evaluation and human feedback workflows. Responsibilities included generating high-quality prompts and benchmark texts, designing detailed evaluation rubrics, and conducting head-to-head comparative assessments of model-generated responses. Performed qualitative and quantitative analysis of outputs based on coherence, factual accuracy, reasoning depth, instruction adherence, and safety compliance. Executed structured question-answering tasks to test domain adaptability and contextual understanding, while documenting edge cases and failure patterns to inform iterative fine-tuning and reinforcement learning cycles.

2025 - 2025

Generalist

Scale AIImageBounding BoxPoint Key Point

Project Scope: Multimodal AI Training & RLHF Optimization Contributed to a multimodal AI development initiative focused on improving Large Language Model (LLM) performance using Reinforcement Learning from Human Feedback (RLHF). Responsibilities included training, annotating, and evaluating model outputs across both text and vision tasks to enhance alignment, accuracy, and generative quality. The task group involved refining workflows for text-to-image generation, image inpainting, object detection and identification, product categorization, entity tagging, and cross-modal response evaluation. Provided structured human feedback, ranked model outputs, identified failure cases, and ensured adherence to quality benchmarks to support iterative model fine-tuning and performance optimization.

Project Scope: Multimodal AI Training & RLHF Optimization Contributed to a multimodal AI development initiative focused on improving Large Language Model (LLM) performance using Reinforcement Learning from Human Feedback (RLHF). Responsibilities included training, annotating, and evaluating model outputs across both text and vision tasks to enhance alignment, accuracy, and generative quality. The task group involved refining workflows for text-to-image generation, image inpainting, object detection and identification, product categorization, entity tagging, and cross-modal response evaluation. Provided structured human feedback, ranked model outputs, identified failure cases, and ensured adherence to quality benchmarks to support iterative model fine-tuning and performance optimization.

2025 - 2025

QA - Generalist

OneformaVideoSegmentationEvaluation Rating

Project Scope: AI Video Annotation – Camera Movement Classification Contributed to a video annotation project focused on training computer vision models to accurately recognize and classify camera movements across documentaries, films, and sports broadcasts. The role involved identifying and labeling camera techniques—such as pan, tilt, dolly, zoom, handheld, tracking, and crane shots—using industry-standard terminology and cinematic conventions. Additionally, conducted quality evaluation and analytical review of annotations submitted by other contributors to ensure consistency, technical accuracy, and adherence to established labeling guidelines. Provided feedback, identified discrepancies, and helped maintain high inter-annotator agreement to improve dataset reliability and model training outcomes.

Project Scope: AI Video Annotation – Camera Movement Classification Contributed to a video annotation project focused on training computer vision models to accurately recognize and classify camera movements across documentaries, films, and sports broadcasts. The role involved identifying and labeling camera techniques—such as pan, tilt, dolly, zoom, handheld, tracking, and crane shots—using industry-standard terminology and cinematic conventions. Additionally, conducted quality evaluation and analytical review of annotations submitted by other contributors to ensure consistency, technical accuracy, and adherence to established labeling guidelines. Provided feedback, identified discrepancies, and helped maintain high inter-annotator agreement to improve dataset reliability and model training outcomes.

2025 - 2025

QA - Generalist

OneformaImageSegmentationEvaluation Rating

Contributed to an AI-driven data labeling initiative focused on optimizing creative generation for digital advertising campaigns. The project involved segregating, evaluating, and classifying AI-generated product images from an e-commerce marketplace based on quality, compliance, visual accuracy, background consistency, branding alignment, and marketplace standards. Key responsibilities included validating image realism, identifying visual artifacts, tagging product categories and attributes, and benchmarking outputs against campaign-ready creative requirements. The labeled dataset was used to train and refine an automated workflow for generating high-performing ad creatives at scale, improving creative turnaround time and reducing manual design intervention.

Contributed to an AI-driven data labeling initiative focused on optimizing creative generation for digital advertising campaigns. The project involved segregating, evaluating, and classifying AI-generated product images from an e-commerce marketplace based on quality, compliance, visual accuracy, background consistency, branding alignment, and marketplace standards. Key responsibilities included validating image realism, identifying visual artifacts, tagging product categories and attributes, and benchmarking outputs against campaign-ready creative requirements. The labeled dataset was used to train and refine an automated workflow for generating high-performing ad creatives at scale, improving creative turnaround time and reducing manual design intervention.

2025 - 2025

Education

Work History