Joe Wentwort

Key Skills

Software

Don't disclose

CVAT

Top Subject Matter

No subject matter listed

Top Data Types

Text

Image

Top Label Types

Evaluation Rating

Classification

Freelancer Overview

I am a Computer Science student with hands-on experience in data labeling, dataset annotation, and reinforcement learning from human feedback (RLHF) for AI and Web3 projects. I have contributed to training large language models by categorizing and labeling complex datasets, ranking AI-generated responses, and ensuring high standards of linguistic nuance and factual accuracy. My technical toolkit includes Google Sheets, Notion, Canva, and basic Python, which I use to manage and evaluate structured data efficiently. With a professional typing speed of 100 WPM and a track record of exceeding quality benchmarks, I am skilled at completing high-volume remote tasks such as transcription, content moderation, and quality control. I am passionate about supporting the development of reliable AI systems through precise data annotation and thoughtful model evaluation.

Labeling Experience

LLM Response Evaluation and Ranking

Don T DiscloseTextEvaluation Rating

Contributed to large scale evaluation of AI generated text responses for reinforcement learning workflows. Ranked multiple model outputs based on factual accuracy, coherence, instruction adherence, tone, and reasoning depth. Applied structured evaluation rubrics to ensure consistent scoring across batches. Reviewed thousands of prompts and responses across diverse domains including general knowledge, reasoning tasks, and conversational AI. Maintained over 95 percent quality compliance based on internal review benchmarks and consistently met daily submission targets.

Contributed to large scale evaluation of AI generated text responses for reinforcement learning workflows. Ranked multiple model outputs based on factual accuracy, coherence, instruction adherence, tone, and reasoning depth. Applied structured evaluation rubrics to ensure consistent scoring across batches. Reviewed thousands of prompts and responses across diverse domains including general knowledge, reasoning tasks, and conversational AI. Maintained over 95 percent quality compliance based on internal review benchmarks and consistently met daily submission targets.

2024 - 2024

Image Annotation and Visual data labelling for computer vision models

CVATImageClassification

Annotated image datasets for supervised computer vision training tasks. Applied bounding boxes and classification labels to objects within images based on predefined annotation guidelines. Ensured accurate object localization, consistent labeling standards, and adherence to dataset taxonomy requirements. Contributed to quality control cycles by reviewing flagged annotations and correcting inconsistencies. Worked on diverse image categories including everyday objects, indoor scenes, and public environment datasets used for object detection and classification models.

Annotated image datasets for supervised computer vision training tasks. Applied bounding boxes and classification labels to objects within images based on predefined annotation guidelines. Ensured accurate object localization, consistent labeling standards, and adherence to dataset taxonomy requirements. Contributed to quality control cycles by reviewing flagged annotations and correcting inconsistencies. Worked on diverse image categories including everyday objects, indoor scenes, and public environment datasets used for object detection and classification models.

2024 - 2025

Education

Work History