Independent Web3 Security Researcher & AI Workflow Specialist (AI coding/model response evaluation and QA-style report generation)
Conducts AI-assisted coding and model response evaluations by assessing correctness, reasoning quality, helpfulness, safety, and instruction compliance for technical workflows. Applies structured evaluation and QA-style reproduction to compare hypotheses, detect reasoning gaps, and improve evidence quality in outputs intended for AI training and coding-evaluation pipelines. Produces clear written evaluations for triage and engineering audiences, with an emphasis on exploit-path clarity and evidence-based reporting. • Assesses model responses and reasoning for quality, gaps, and instruction following • Compares AI/model outputs while investigating unfamiliar or complex code behavior • Builds and reviews reproduction materials using Foundry, Hardhat, and Clarinet where applicable • Writes structured, reviewable technical reports using PoC-driven and expected-vs-actual style evidence