Contract AI Safety Researcher & Red Teamer
Selected for model safety work; conducted adversarial testing and identified universal jailbreak-style vulnerabilities in sensitive domains; contributed to mitigations pre-deployment.
04 / Service & work history
Selected for model safety work; conducted adversarial testing and identified universal jailbreak-style vulnerabilities in sensitive domains; contributed to mitigations pre-deployment.
Authored and published NLP research (“Town Hall Debate Prompting”) and continue building research workflows for evaluation-driven iteration.
Engaged in philosophical discussions to train AI in navigating complex conversations. Assisted in evaluating AI's performance and improving communication in polarizing discourse.
Assisted in cold outreach and data analysis, sending over 100 cold outreach emails and messages to secure 60+ survey responses.
Completed a 5-week fellowship, working on 5 AI projects, culminating in a final showcase project. Developed web development skills and product creation.
Experience
Direct a grant-funded, news-featured community initiative: organize events, secure funding, and manage outreach. 500+ documented service hours.
Engaged in a district-wide music nonprofit with external funding and media coverage. 100+ additional documented service hours through performances and events.
Captain of a #1 NSDA-ranked program. Under my captaincy: Mundt Sweepstakes Trophy, Harvard Classic sweep winner, TOC/Yale sweeps (2nd), and national-level placements.
Co-founded a competition-focused education nonprofit; curriculum emphasizes derivations, proofs, practice, and conceptual mastery. Sponsors include Jane Street, Art of Problem Solving, and Wolfram Research.
Education
Palm Beach Campus · Delray Beach, FL · Class of 2027
Capabilities