Vienna / AI Systems

Loren Angoni

AI Systems and Governance Architect, Vienna. Building runtime governance for AI agents in regulated environments. Open-source: fivedrisk.

My Writings About me

Featured Writing

Systems, governance, and the decisions in between.

June 13, 2026

Runtime Is the Moment of Consequence in AI

A radiation machine killed three people because of operator fatigue, training that said failure was impossible, and no safety layer at the moment it acted. Today's AI agents are running the same pattern.

16 min read runtime-governance / regulation / historical-analogues

May 11, 2026

AI Governance Without Runtime Is Performative

Most AI governance does not constrain agent behavior at runtime. The runtime layer is the missing piece for deploying agentic AI at scale in high-risk or regulated environments.

9 min read runtime-governance / regulation

April 11, 2026

AI Agent Approval Fatigue Is Real

If your coding agent asks for approval on every harmless action, people either turn it off or stop using the loop. The fix is not more prompts. It is a policy layer.

2 min read runtime-governance / ai-architecture

What I’m Exploring

5D and DotOS: runtime policy and the systems behind it.

I’m building a runtime policy engine for AI agents and the larger systems lab around it: approvals, tool boundaries, governance, memory, and what architecture makes longer autonomous loops trustworthy enough to use.

See the projects