Radji Moustapha

Key Skills

Software

AWS SageMaker

Appen

CVAT

Diffgram

Labelbox

LabelImg

Label Studio

OpenCV AI Kit (OAK)

Prodigy

Remotasks

Scale AI

SuperAnnotate

Supervisely

Toloka

Top Subject Matter

No subject matter listed

Top Data Types

Audio

Image

Text

Top Task Types

Bounding Box

Classification

Entity Ner Classification

Object Detection

Segmentation

Freelancer Overview

Having worked on the ground for more than 5 years in data labeling and AI training data workflows, I have been involved in game-changing projects in image, text, and audio modalities. My areas of expertise include a wide variety of annotation tools such as CVAT, Labelbox, LabelImg, Label Studio, Remotasks, AWS SageMaker, and in-house platforms. I have helped machine learning teams by conducting accurate and high-quality annotation for computer vision models (bounding boxes, segmentation, object detection), natural language processing tasks (NER, classification, summarization), and multilingual audio transcription.I offer intense attention to detail, focus on strict quality controls, and the ability to adjust to new labeling environments and toolsets in no time. My past experience includes work in various fields ranging from autonomous driving and e-commerce to healthcare and LLM evaluation, which makes me a reliable member of a large variety of AI training pipelines.

Labeling Experience

Healthcare Image Annotation for Medical AI Models

Label StudioImageSegmentationClassification

Was a data annotator for a medical AI project, which was to automate the process of detecting tumors and abnormalities in MRI and CT scan images. Annotated over 10,000 medical images, pixel-level segmentation, and lesion and tumor classification. Worked with medical practitioners to ensure that the medical conditions like cancerous cells, benign growths, and others were correctly labeled. The project strictly followed the data privacy rules and used specific software solutions, such as Label Studio and Dataloop, for uniform annotations. Have helped to create a deep learning model that attained 95% accuracy in the identification of important medical features in diagnostic images.

Was a data annotator for a medical AI project, which was to automate the process of detecting tumors and abnormalities in MRI and CT scan images. Annotated over 10,000 medical images, pixel-level segmentation, and lesion and tumor classification. Worked with medical practitioners to ensure that the medical conditions like cancerous cells, benign growths, and others were correctly labeled. The project strictly followed the data privacy rules and used specific software solutions, such as Label Studio and Dataloop, for uniform annotations. Have helped to create a deep learning model that attained 95% accuracy in the identification of important medical features in diagnostic images.

2021

Autonomous Vehicle Object Detection Annotation Project

CVATImageBounding BoxSegmentation

Worked on a large-scale annotation project for a self-driving car company, focusing on object detection and classification across urban and suburban traffic scenarios. Labeled over 50,000 high-resolution street-level images with bounding boxes and segmentation masks for vehicles, pedestrians, road signs, and lane markings. Used CVAT and Labelbox extensively to ensure consistency, accuracy, and quality. Applied internal QA guidelines and cross-reviewed samples with other annotators to maintain over 98% accuracy. Participated in weekly review cycles with AI engineers to improve annotation standards and resolve ambiguity in edge cases.

Worked on a large-scale annotation project for a self-driving car company, focusing on object detection and classification across urban and suburban traffic scenarios. Labeled over 50,000 high-resolution street-level images with bounding boxes and segmentation masks for vehicles, pedestrians, road signs, and lane markings. Used CVAT and Labelbox extensively to ensure consistency, accuracy, and quality. Applied internal QA guidelines and cross-reviewed samples with other annotators to maintain over 98% accuracy. Participated in weekly review cycles with AI engineers to improve annotation standards and resolve ambiguity in edge cases.

2020 - 2022

Education

Work History