Propio

Design and maintain safety evaluation frameworks and lead red-teaming exercises to identify vulnerabilities in LLMs and agents. Build runtime guardrails and implement threat modeling to ensure agents operate within secure, well-defined boundaries.