Matthew Rosenfeld

Key Skills

Software

Scale AI

Top Subject Matter

No subject matter listed

Top Data Types

Image

Text

Video

Top Label Types

Bounding Box

Classification

Prompt Response Writing SFT

Segmentation

Transcription

Freelancer Overview

I am an experienced AI data annotator and data labeling specialist with a strong background in computer science and hands-on expertise supporting machine learning projects. I have worked extensively on annotating large-scale image, text, and audio datasets, applying detailed guidelines with high accuracy and consistency. My experience spans computer vision tasks such as object detection and segmentation, as well as NLP annotation for sentiment analysis and text classification. I am skilled in using industry-standard tools like Labelbox, CVAT, SuperAnnotate, and have a working knowledge of Python, Excel, and Google Sheets. I take pride in maintaining a 98%+ annotation accuracy rate and am recognized for my attention to detail, reliability, and ability to quickly adapt to new annotation frameworks. My goal is to contribute to high-quality AI development by delivering precise and reliable training data for innovative projects.

Labeling Experience

AI Data Annotator

Scale AIImageBounding BoxSegmentation

Contributed to a large-scale image annotation project supporting AI and machine learning model development. The project involved labeling thousands of images using Remotasks, applying both bounding box annotation to detect and localize objects and segmentation techniques to capture precise pixel-level boundaries for complex regions and shapes. Tasks required strict adherence to detailed annotation guidelines to ensure consistency and accuracy across the dataset. Maintained a 98%+ accuracy rate through regular quality assurance checks and self-review processes, contributing to a reliable, high-quality training dataset used to improve computer vision model performance.

Contributed to a large-scale image annotation project supporting AI and machine learning model development. The project involved labeling thousands of images using Remotasks, applying both bounding box annotation to detect and localize objects and segmentation techniques to capture precise pixel-level boundaries for complex regions and shapes. Tasks required strict adherence to detailed annotation guidelines to ensure consistency and accuracy across the dataset. Maintained a 98%+ accuracy rate through regular quality assurance checks and self-review processes, contributing to a reliable, high-quality training dataset used to improve computer vision model performance.

2020 - 2024

Education

Work History