Steele H.

Key Skills

Software

No software listed

Top Subject Matter

AI tutoring/training data preparation

preference/model alignment

safety evaluation

Top Data Types

Text

Image

Top Task Types

Entity (NER) Classification

RLHF

Red Teaming

Freelancer Overview

AI Trainer / Data Annotation Specialist, MindSpec. Brings 6+ years of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include Internal and Proprietary Tooling. Education includes Master of Science, University of Florida (2021) and Bachelor of Science, University of Central Florida (2019). AI-training focus includes data types such as Text and labeling workflows including Evaluation, Rating, and Named Entity Recognition.

Labeling Experience

AI Trainer / Data Annotation Specialist, MindSpec

TextEntity (NER) Classification

Served as an AI Trainer and data annotation specialist by guiding annotators and improving labeling performance for production AI systems. Coordinated calibration and quality assurance activities to systematically increase label accuracy. Collaborated with ML engineers to translate observed model failures into targeted data collection and relabeling strategies. • Trained and calibrated 8 junior annotators, improving average label accuracy from 87% to 96% • Built feedback loops from model failure modes to drive targeted relabeling strategies • Conducted manual red teaming and adversarial prompt testing for harmful/bias/quality detection • Supported prompt-response evaluation processes to improve model alignment prior to deployment

2023 - 2026

Data Annotation Associate, BlackSky

TextRLHFRed Teaming

Worked as a data annotation associate supporting CV and NLP client projects through dataset tagging and validation. Completed both labeling and quality checks to ensure consistency and throughput in production. Contributed to RLHF pilot pipelines by ranking chatbot responses for helpfulness, harmlessness, and factual accuracy. • Tagged and validated text, image, and audio datasets for CV and NLP models • Maintained personal annotation accuracy of 98% while meeting throughput targets • Ranked chatbot responses for helpfulness, harmlessness, and factual accuracy (RLHF) • Reported systemic tool issues and provided input for workflow improvements

2021 - 2022

Education

Work History