Reda Taraoui

Key Skills

Software

AWS SageMaker

Diffgram

Labelbox

SuperAnnotate

Top Subject Matter

No subject matter listed

Top Data Types

Image

Text

Video

Top Task Types

Action Recognition

Bounding Box

Classification

Fine Tuning

Object Detection

Freelancer Overview

Bilingual (English/French) AI Data Labeler and Annotation Specialist with hands-on experience preparing, annotating, and validating datasets for AI/ML training. Skilled in text and voice annotation, transcription, and prompt-response evaluation for generative AI systems. At Capgemini (Biosphere), I labeled and validated soil and environmental datasets using SuperAnnotate and Amazon SageMaker Ground Truth, contributing to accurate training data for ML models. I also supported data quality validation and automation workflows using SQL, Airflow, and BigQuery. Known for precision, attention to detail, and adaptability across text, audio, and scientific data projects, I am motivated to contribute high-quality labeled data that improves the reliability, safety, and fairness of AI systems.

Reviews (1)

Labeling Experience

Text & Voice Annotation (Capgemini + Freelance Support)

SuperannotateTextEntity Ner ClassificationClassification

Conducted bilingual (English & French) text and voice annotation tasks, including transcription, entity recognition, and sentiment/intent classification. Evaluated and refined AI-generated text responses to improve accuracy, tone, and relevance in large language models. Delivered thousands of high-quality annotations under strict project guidelines, maintaining accuracy above 97%. Contributed structured feedback on LLM behavior to support fine-tuning and safety evaluation.

Conducted bilingual (English & French) text and voice annotation tasks, including transcription, entity recognition, and sentiment/intent classification. Evaluated and refined AI-generated text responses to improve accuracy, tone, and relevance in large language models. Delivered thousands of high-quality annotations under strict project guidelines, maintaining accuracy above 97%. Contributed structured feedback on LLM behavior to support fine-tuning and safety evaluation.

2024 - 2025

Soil Dataset Annotation (Capgemini Biosphere)

Aws SagemakerImageBounding BoxPolygon

Annotated and validated large-scale soil and environmental datasets to create high-quality training data for machine learning models. Tasks included drawing bounding boxes and polygons around soil samples, tagging categories, and ensuring consistent labeling according to project guidelines. Delivered thousands of annotations with high accuracy, supporting model training and evaluation. Applied strict quality checks and collaborated with data scientists to refine labeling protocols.

Annotated and validated large-scale soil and environmental datasets to create high-quality training data for machine learning models. Tasks included drawing bounding boxes and polygons around soil samples, tagging categories, and ensuring consistent labeling according to project guidelines. Delivered thousands of annotations with high accuracy, supporting model training and evaluation. Applied strict quality checks and collaborated with data scientists to refine labeling protocols.

2024 - 2025

Education

Work History