Khaled Hamed

Key Skills

Software

Label Studio

Appen

Scale AI

CVAT

AWS SageMaker

Top Subject Matter

No subject matter listed

Top Data Types

Computer Code Programming

Geospatial Tiled Imagery

Medical Dicom

Text

Video

Top Label Types

Classification

Entity Ner Classification

Freelancer Overview

I’m an Oracle-certified Data Scientist with a strong foundation in machine learning, data quality, and structured problem-solving. I work comfortably with Python and SQL, and I’m used to turning messy inputs into clean, consistent datasets through validation, error checking, and clear documentation. I learn new taxonomies quickly and I’m detail-oriented when applying rules consistently—especially when requirements are strict. While I’m newer to paid AI data labeling, my data science background (and medical studies) makes me effective on projects that require careful reading, domain awareness, and high-precision decisions. I’m confident working with text classification tasks, building/using annotation guidelines, tracking edge cases, and communicating clearly with reviewers to improve consistency and quality over time.

Labeling Experience

AI Data Labeler / AI Trainer (Freelance)

Label StudioTextClassification

Provided expert-level data labeling and classification for diverse text and multimodal datasets. Applied guideline-based annotation and rubric-driven evaluation for LLM outputs, emphasizing labeling quality and error reduction. Regularly performed consensus QA, spot checks, and error analysis, with a strong focus on healthcare and biomedical subject matter. • Labeled and evaluated datasets as a freelancer on Khamsat.com and other platforms. • Used Label Studio, CVAT, Argilla, and spreadsheets for high-fidelity annotation and QA. • Supported LLM rubric scoring, preference ranking, and safety/policy labeling. • Specialized in medical terminology and biomedical text labeling.

Provided expert-level data labeling and classification for diverse text and multimodal datasets. Applied guideline-based annotation and rubric-driven evaluation for LLM outputs, emphasizing labeling quality and error reduction. Regularly performed consensus QA, spot checks, and error analysis, with a strong focus on healthcare and biomedical subject matter. • Labeled and evaluated datasets as a freelancer on Khamsat.com and other platforms. • Used Label Studio, CVAT, Argilla, and spreadsheets for high-fidelity annotation and QA. • Supported LLM rubric scoring, preference ranking, and safety/policy labeling. • Specialized in medical terminology and biomedical text labeling.

2018

Product Identification & Shelf Auditing for Retail AI

Aws SagemakerImagePolygon

Worked on a large-scale data labeling project to train a computer vision model for automated retail inventory management. My primary responsibility was to accurately identify and label 150+ different types of consumer goods (SKUs) on grocery store shelves using 2D Bounding Boxes. Key responsibilities included: Annotating over 10,000 high-resolution images with sub-pixel precision. Differentiating between similar-looking products (e.g., different flavors of the same brand). Handling "occlusion" cases where products were partially hidden or blurry. Adhering to strict quality control measures, consistently achieving a 99% accuracy rate across internal "Golden Set" audits.

Worked on a large-scale data labeling project to train a computer vision model for automated retail inventory management. My primary responsibility was to accurately identify and label 150+ different types of consumer goods (SKUs) on grocery store shelves using 2D Bounding Boxes. Key responsibilities included: Annotating over 10,000 high-resolution images with sub-pixel precision. Differentiating between similar-looking products (e.g., different flavors of the same brand). Handling "occlusion" cases where products were partially hidden or blurry. Adhering to strict quality control measures, consistently achieving a 99% accuracy rate across internal "Golden Set" audits.

2025 - 2025

3D Cuboid and Semantic Segmentation for Autonomous Driving

CVATVideoBounding BoxSegmentation

Performed high-precision annotation of LIDAR and camera data for a Level 4 autonomous driving system. The scope involved identifying and labeling dynamic objects (pedestrians, cyclists, vehicles) using 3D Cuboids and Semantic Segmentation for road boundaries. I handled complex "occlusion" scenarios where objects were partially hidden. Consistently met a throughput of 20+ complex frames per hour while maintaining a sub-pixel accuracy requirement for safety-critical data.

Performed high-precision annotation of LIDAR and camera data for a Level 4 autonomous driving system. The scope involved identifying and labeling dynamic objects (pedestrians, cyclists, vehicles) using 3D Cuboids and Semantic Segmentation for road boundaries. I handled complex "occlusion" scenarios where objects were partially hidden. Consistently met a throughput of 20+ complex frames per hour while maintaining a sub-pixel accuracy requirement for safety-critical data.

2025 - 2025

RLHF Evaluation and Prompt Engineering for Large Language Models

Scale AITextRLHFFine Tuning

Collaborated on a large-scale project to improve the reasoning and safety of a frontier LLM. Tasks included ranking multiple AI responses based on truthfulness, helpfulness, and harmlessness (RLHF). I authored high-complexity prompts to test the model's edge cases in coding (Python/JS) and creative writing. Maintained a 98% quality score across 500+ tasks, adhering to strict multi-step reasoning guidelines and fact-checking protocols.

Collaborated on a large-scale project to improve the reasoning and safety of a frontier LLM. Tasks included ranking multiple AI responses based on truthfulness, helpfulness, and harmlessness (RLHF). I authored high-complexity prompts to test the model's edge cases in coding (Python/JS) and creative writing. Maintained a 98% quality score across 500+ tasks, adhering to strict multi-step reasoning guidelines and fact-checking protocols.

2025 - 2025

AI Financial Advisor: Intent Classification, Function-Calling, and SQL Validation

Aws SagemakerTextClassificationEvaluation Rating

Worked on a high-complexity dataset for training an AI-driven Personal Finance Management (PFM) advisor. My role involved multi-layered annotation of financial conversations to improve the model's reasoning and utility. Key Responsibilities: Intent & Category Classification: Labeled user queries for intent (e.g., "spending insights," "budgeting") and categorized transactions/merchants with high accuracy. Function-Call Tagging: Identified and tagged backend data queries (e.g., get_user_summary) for API-style structured data, ensuring the model triggered the correct software functions. SQL Safety & Validation: Evaluated and labeled custom SQL queries to ensure they were "safe," "parameterized," and "read-only," preventing potential security risks. Response Evaluation: Rated AI assistant responses based on tone, professionalism, and completeness. Structured Output: Delivered final datasets in structured JSON format, maintaining strict data integrity and financial terminology standards.

Worked on a high-complexity dataset for training an AI-driven Personal Finance Management (PFM) advisor. My role involved multi-layered annotation of financial conversations to improve the model's reasoning and utility. Key Responsibilities: Intent & Category Classification: Labeled user queries for intent (e.g., "spending insights," "budgeting") and categorized transactions/merchants with high accuracy. Function-Call Tagging: Identified and tagged backend data queries (e.g., get_user_summary) for API-style structured data, ensuring the model triggered the correct software functions. SQL Safety & Validation: Evaluated and labeled custom SQL queries to ensure they were "safe," "parameterized," and "read-only," preventing potential security risks. Response Evaluation: Rated AI assistant responses based on tone, professionalism, and completeness. Structured Output: Delivered final datasets in structured JSON format, maintaining strict data integrity and financial terminology standards.

2024 - 2025

Show more

Education

Work History