For employers

Hire this AI Trainer

Sign in or create an account to invite AI Trainers to your job.

Invite to Job
P
Pure G.

Pure G.

Full Stack Developer in Research, Information Technology, and Mathematics

USA flagMaryland, Usa

Key Skills

Software

Other

Top Subject Matter

General Research
Information Technology
Mathematics

Top Data Types

TextText
DocumentDocument
ImageImage

Top Task Types

Evaluation/RatingEvaluation/Rating
Text GenerationText Generation
Question AnsweringQuestion Answering

Freelancer Overview

I’m still pretty new on paper when it comes to AI training data / labeling work—I’ve spent a few weeks total on mostly rating/evaluating model responses and doing factuality / quality checks against provided criteria. I liked the work because it rewards solid comprehension and consistent judgment—and noticing mistakes that could otherwise slip through the cracks. Most of my relevant muscle comes from elsewhere: I’ve spent hundreds if not thousands of hours working on a real internal product where getting the data “right” matters more than shipping slop: a Python/Streamlit app with Google Sheets/Drive integrations and SQLite/Turso-backed workflows that replaces a lot of manual monthly finance work across revenue, allocations, postings, and landlord-facing outputs. A big part of the job was translating what people meant operationally into concrete rules, then validating results against the firm’s existing spreadsheet practice until the numbers lined up—basically the same habits you need in annotation work: interpret instructions, catch edge cases early, and tighten things through feedback. Before that, I worked in fire protection inspection and maintenance, which taught me to document what I did, follow structured processes, and communicate clearly. I pick up new tools quickly when I’m given a real problem to solve—I learn by doing rather than collecting certificates—so I always love the opportunity to be evaluated on consistency and quality rather than on how long I’ve been in the field.

Labeling Experience

LLM response evaluation & factuality

TextTextEvaluation/RatingEvaluation/Rating

Small contract piece: judge prompt + two AI responses. Main tasks were rubric scoring, checking whether answers matched facts and details (lots of lightweight research/verification), and following the evaluator guidelines consistently. Scope was modest—repeatable workflows, unpredictable content. Often, one answer was nearly there while the other had obvious factual or logic problems—so diligence mattered.

2026 - 2026

Education

U

University of Maryland, Shady Grove

Bachelor of Science, Information Science

Bachelor of Science
2020 - 2022
A

Anne Arundel Community College, Arnold

Associate of Arts, Transfer Studies

Associate of Arts
2017 - 2019

Work History

C

Corner Media

Full Stack Developer

Virginia
2025 - Present
A

Anne Arundel Fire Protection

Fire Systems Inspector

Edgewater
2024 - 2025