David Quispe

Hi, I’m a systems engineer working across cybersecurity and AI safety.

I build practical AI systems with an emphasis on agent behavior: agentic workflows, synthetic data pipelines, LLM evaluation loops, and secure deployments with planning, tool use, and auditable execution. My work sits where model behavior, safety constraints, product judgment, and operational risk meet.

In cybersecurity, I focus on vulnerability intelligence, secure architecture, adversarial threat modeling, and compliance-aware controls across hybrid environments. I work with frameworks such as ISO/IEC 27001/22301, NIST CSF, ENS, GDPR, and the EU AI Act.

Right now I am focused on PERSONA.md and Personaxis: an open behavioral specification and registry/deployment platform for AI agent personas. The goal is to make the full persona structure portable, versionable, auditable, and safer to govern across tools, teams, and model updates: identity is only one layer among character, personality, cognition, affect, drives and values, normative self-regulation, memory, metacognition, and presentation.