Henry Derick

I’m a detail-oriented AI Data Annotator and Evaluator, bringing specialized domain expertise across Health Services Management, STEM, and Cybersecurity. I evaluate and rate Large Language Model (LLM) responses for factual accuracy, logical consistency, and safety alignment, with a 99.8% quality score across 10,000+ completed tasks. I’m equally comfortable stress-testing reasoning through complex, multi-turn prompt work.

My core strength is making AI outputs dependable: I perform Reinforcement Learning from Human Feedback (RLHF) and deliver highly detailed written feedback to engineers on hallucination rates and contextual failures. Using Open Source Intelligence (OSINT), I complete rigorous fact-checking for obscure data points and technical claims—reducing model hallucination rates by over 40% in targeted testing domains. I also verify academic-style logic and technical reasoning to support high-precision evaluation of advanced reasoning models.

Alongside AI evaluation, I work as a Cybersecurity & Network Analyst (independent/self-directed), analyzing network traffic, security logs, and malware behaviors. I use Wireshark, Nmap, Burp Suite, and Kali Linux, and complete rigorous vulnerability assessment labs via PortSwigger and TryHackMe, simulating over 50 advanced red-team engagements. I’m pursuing a Bachelor’s Degree in Health Services Management (expected 2026), which strengthens my ability to label and evaluate domain-specific medical/health logic.