Daeshauwna Jones

Key Skills

Software

AWS SageMaker

Appen

Clickworker

CloudFactory

CrowdFlower

Data Annotation Tech

Figure Eight

Labelbox

Label Studio

Lionbridge

Mindrift

Remotasks

Scale AI

Snorkel AI

SuperAnnotate

Toloka

Telus

Internal/Proprietary Tooling

Top Subject Matter

No subject matter listed

Top Data Types

Audio

Medical Dicom

Text

Top Task Types

Bounding Box

Classification

Evaluation Rating

Prompt Response Writing SFT

Question Answering

Freelancer Overview

I'm a seasoned AI Data Labeling and NLP Specialist with over 7 years of experience supporting top-tier clients through platforms like Upwork, Appen, Toloka, and Scale AI. I’ve contributed to diverse projects spanning LLM evaluation, multilingual named entity recognition (NER), image segmentation for medical diagnostics, sentiment classification, and prompt engineering for fine-tuning AI systems. My work stands out for its high accuracy, fast turnaround, and strong quality assurance scores. I specialize in evaluating large language model outputs (LLMs) for safety, relevance, and factual integrity, as well as crafting complex prompt-response pairs across multiple languages. I’m also skilled in tools like Label Studio, CVAT, and AWS SageMaker. With a background in linguistics and a sharp eye for detail, I bring consistency, adaptability, and insight to every project I take on.

Labeling Experience

LLM Prompt Evaluation & Safety Rating

Scale AITextClassificationQuestion Answering

Evaluated thousands of AI-generated responses for safety, helpfulness, coherence, tone, and factual accuracy using proprietary guidelines. Tasks included rating LLM outputs, identifying harmful or biased content, performing red teaming, and writing prompt-response pairs for fine-tuning. This project supported multilingual evaluation (English, Spanish) and required attention to linguistic nuance and ethical alignment. Delivered consistent, high-quality feedback with a QA approval rate exceeding 95%. Labeled over 500 hours of training and validation data in collaboration with global AI teams.

Evaluated thousands of AI-generated responses for safety, helpfulness, coherence, tone, and factual accuracy using proprietary guidelines. Tasks included rating LLM outputs, identifying harmful or biased content, performing red teaming, and writing prompt-response pairs for fine-tuning. This project supported multilingual evaluation (English, Spanish) and required attention to linguistic nuance and ethical alignment. Delivered consistent, high-quality feedback with a QA approval rate exceeding 95%. Labeled over 500 hours of training and validation data in collaboration with global AI teams.

2022 - 2023

Medical Image Segmentation & Classification for Diagnostic AI

Scale AIImageBounding BoxSegmentation

Performed pixel-level image segmentation and classification of radiology scans including X-rays and ultrasound images. Used CVAT and Scale AI platforms to label anatomical structures, highlight regions of interest, and identify potential abnormalities (e.g., lesions, fluid buildup). Collaborated closely with QA reviewers to ensure high clinical accuracy and consistency. The dataset was used to train and validate early-stage diagnostic AI models aimed at supporting radiologists and improving early disease detection. Maintained 97%+ accuracy across more than 20,000 labeled images.

Performed pixel-level image segmentation and classification of radiology scans including X-rays and ultrasound images. Used CVAT and Scale AI platforms to label anatomical structures, highlight regions of interest, and identify potential abnormalities (e.g., lesions, fluid buildup). Collaborated closely with QA reviewers to ensure high clinical accuracy and consistency. The dataset was used to train and validate early-stage diagnostic AI models aimed at supporting radiologists and improving early disease detection. Maintained 97%+ accuracy across more than 20,000 labeled images.

2019 - 2021

Education

Work History