Bratton C.

Key Skills

Software

Scale AI

Micro1

Top Subject Matter

Large language model evaluation and customer support AI training.

Customer support conversation training

Ranking Domain Expertise

Top Data Types

Text

Document

Top Task Types

RLHF

Freelancer Overview

Outlier AI | AI Generalist Expert | Freelance AI Trainer Project. Brings 7+ years of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include Scale AI and Micro1. Education includes Bachelor of Science, University of South Carolina. AI-training focus includes data types such as Text and labeling workflows including Evaluation, Rating, and RLHF.

Labeling Experience

Micro1 | Customer Support AI Trainer Specialist

Micro1TextRLHF

Trained and evaluated AI systems for customer support by reviewing model outputs for empathy, clarity, professionalism, issue resolution, and contextual accuracy. Created and refined customer support prompts to teach handling of troubleshooting, escalations, refunds, onboarding, and live communication scenarios. Reviewed, ranked, and used qualitative metrics such as tone, helpfulness, compliance, and response quality to reduce low-quality outputs through iterative feedback loops. • Reviewed and ranked 3,000+ customer support conversations based on response quality metrics. • Reduced low-quality customer support responses by 22% via prompt optimization and evaluation workflows. • Performed structured feedback and evaluation aligned with RLHF-style improvement processes. • Supported ongoing AI training for customer interaction use cases.

2025 - Present

Outlier AI | AI Generalist Expert | Freelance AI Trainer Project

Scale AIText

Performed AI response evaluations for customer support and LLM training projects by scoring quality dimensions including reasoning, factuality, coherence, and instruction-following. Produced structured rationales and scoring feedback aligned with RLHF-style reinforcement learning workflows. Conducted side-by-side comparisons to identify hallucinations, logical inconsistencies, tone issues, and factual inaccuracies for iterative model improvement. • Completed 4,200+ prompt-response evaluations and annotation tasks with sustained 98% QA accuracy. • Applied rubric-based feedback to refine AI behavior and conversational quality. • Assisted with large language model output optimization for customer support scenarios. • Evaluated responses across multiple AI initiatives using QA frameworks.

2024 - Present

Education

Work History