Mordecai Machazire

AI systems architect and researcher at Project Hamburg Research exploring one question: how do we make AI trustworthy? I build and study governed agent systems, zero-trust AI architecture, agent evaluation, behavioral observability, and methods for helping agents remain grounded in their task, environment, and constraints. My work includes SURLG, curriculum-based agent onboarding, AI visibility research, and emerging approaches to multi-agent governance and attunement. MPH-trained in epidemiology and biostatistics; AI-native builder working across research, security, and agentic systems.

• AI Governance
• AI Safety
• Agentic AI
• Multi-Agent Systems
• Agent Evaluation
• LLM Evaluation
• AI Security
• Zero-Trust Architecture
• Agent Observability
• Behavioral Evaluation
• Adversarial Testing
• RAG
• AI Visibility
• Rubric Design
• Research Systems
• LLM APIs
• Python
• SQL
• Epidemiology
• Biostatistics