Writing

Writing.

I write about AI governance, agent deployment, trust infrastructure, and what it actually looks like to build a company in this space.

Planned Posts

• Why I Started VeriSwarm — The founder-origin story behind building trust and security infrastructure for AI agents.
  Suggested slug: /writing/why-i-started-veriswarm
• What Infrastructure Engineers See in AI That App Builders Often Miss — A systems view on reliability, control, and the difference between demos and production.
  Suggested slug: /writing/what-infrastructure-engineers-see-in-ai
• AI Agents Are Being Treated Like Features. They Should Be Treated Like Operators. — A thesis-driven post about why agent systems need operational discipline, not just product polish.
  Suggested slug: /writing/ai-agents-should-be-treated-like-operators
• The Trust Problem Comes Before the Intelligence Problem — Why agent capability matters less than whether the system can be inspected, governed, and trusted.
  Suggested slug: /writing/the-trust-problem-comes-before-the-intelligence-problem
• What Regulated Industries Actually Need Before They Deploy AI Agents — The controls, auditability, and operational guarantees missing from most agent stacks.
  Suggested slug: /writing/what-regulated-industries-need-before-deploying-ai-agents

Additional Topics

• The Missing Layer in Most Agent Stacks — Identity, policy, audit, and evidence as core infrastructure rather than product add-ons.
  Suggested slug: /writing/the-missing-layer-in-most-agent-stacks
• Why Prompt Injection Is an Infrastructure Problem, Not Just a Model Problem — A founder/operator companion to the VeriSwarm argument about tool-path security.
  Suggested slug: /writing/prompt-injection-is-an-infrastructure-problem
• MCP Adoption Is Moving Faster Than MCP Security — A founder-level take on the gap between protocol adoption and operational safeguards.
  Suggested slug: /writing/mcp-adoption-is-moving-faster-than-mcp-security
• The Difference Between Logging and Evidence — Why ordinary observability is not enough for high-trust or regulated environments.
  Suggested slug: /writing/the-difference-between-logging-and-evidence
• Why Just Add a Kill Switch Is Not a Safety Strategy — A more opinionated companion to the operational shutdown problem in agent systems.
  Suggested slug: /writing/why-a-kill-switch-is-not-a-safety-strategy
• What I Learned Running Infrastructure Where Downtime Had Real Consequences — A post grounded in production accountability, operational risk, and what that changes about AI deployment thinking.
  Suggested slug: /writing/what-i-learned-running-infrastructure-where-downtime-had-real-consequences
• The Case for Trust Tiers in Agent Systems — Why different agents should earn different levels of autonomy and access over time.
  Suggested slug: /writing/the-case-for-trust-tiers-in-agent-systems
• Why AI Governance Fails When It Lives Only in Policy Documents — A direct argument for moving governance into systems, controls, and enforcement paths.
  Suggested slug: /writing/why-ai-governance-fails-when-it-lives-only-in-policy-documents
• Engineers, Not Committees, Will Decide What Responsible AI Deployment Actually Looks Like — A strong opinion piece about where real accountability will come from.
  Suggested slug: /writing/engineers-not-committees-will-decide-responsible-ai-deployment
• How I Think About Identity for Non-Human Actors — A direct statement of how identity should work for agents, automations, and delegated software actors.
  Suggested slug: /writing/how-i-think-about-identity-for-non-human-actors

Related Company Writing

Several of these essays will complement existing VeriSwarm posts on trust scoring, MCP security, prompt injection, EU AI Act compliance, and incident-response controls for AI systems.

For shorter takes in the meantime, you can find me on LinkedIn.