Founder · AI Agents · Security
AI Agents are becoming the new browsers.
I build and secure what runs behind them.
Founder & CEO at Jutsu
Author of "MCP Security"
Current Focus
→
AgentSOC: AI-driven SOC automation using multi-agent systems (triage, enrichment, response)
→
MCP Security: A practical guide to the new AI agent attack surface
→
Jutsu Platform: Deploy, observe, and secure agentic systems
The problem I'm working on
AI agents don't fail like software. They fail like distributed systems with intent.
- •MCP creates invisible trust boundaries
- •Tools become implicit privileges
- •Prompt injection becomes lateral movement
- •SOC workflows weren't designed for agents
Mental Models I Use
The Agent Perimeter Model
How agent capabilities expand beyond identity
The MCP Threat Ladder
From: prompt injection → tool abuse → escalation
Self-Healing Agent Loop
Detect → checkpoint → recover → continue
Work with me
I work with a small number of teams building or securing AI agents in production.
If this resonates, email me: