Henry Esu

Key Skills

Software

Data Annotation Tech

Top Subject Matter

No subject matter listed

Top Data Types

Text

Top Label Types

Computer Programming Coding

Evaluation Rating

RLHF

Freelancer Overview

I am a Computer Science student with hands-on experience as an AI Data Researcher, specializing in data curation, annotation, and reinforcement learning from human feedback (RLHF) for large language models. My work involves meticulously reviewing and fact-checking AI-generated content, identifying hallucinations, bias, and safety risks to ensure outputs are accurate and align with human reasoning. I excel at creating complex prompts to test model robustness, ranking and providing detailed feedback on AI responses, and collaborating with technical teams to improve training data quality. My strong background in logic and research allows me to analyze step-by-step reasoning in AI outputs, making me passionate about building reliable and trustworthy AI systems.

Labeling Experience

AI Data Researcher

Data Annotation TechTextRLHFEvaluation Rating

As an AI Data Researcher, I focused on optimizing LLM performance through meticulous data annotation, RLHF, and evaluation. I specialized in identifying model hallucinations, verifying factual accuracy, and assessing reasoning in generated responses. My work included sentiment analysis, entity linking, and evaluation of AI outputs for logic and flow. • Evaluated 1,000+ AI-generated responses for accuracy, logic, and safety. • Performed deep-dive fact-checking and source verification for data integrity. • Conducted prompt engineering and strict adherence to annotation guidelines. • Specialized in hallucination detection and ranking AI responses for trustworthiness.

As an AI Data Researcher, I focused on optimizing LLM performance through meticulous data annotation, RLHF, and evaluation. I specialized in identifying model hallucinations, verifying factual accuracy, and assessing reasoning in generated responses. My work included sentiment analysis, entity linking, and evaluation of AI outputs for logic and flow. • Evaluated 1,000+ AI-generated responses for accuracy, logic, and safety. • Performed deep-dive fact-checking and source verification for data integrity. • Conducted prompt engineering and strict adherence to annotation guidelines. • Specialized in hallucination detection and ranking AI responses for trustworthiness.

2025

Education

Work History