Portfolio & Insights

Open-source agentic AI projects, a seven-part governance white-paper series, case studies, and two US patents — from 35+ years building production systems.

Agentic AI Governance in Practice

Read the series →

A seven-part white-paper series that takes an agentic AI system from governance concept to monitored production. Start with the delivery workflow, which ties the other six together.

  1. Start here · Part 7 Agentic Delivery Workflow v1.10 An eight-stage, gate-closed path from request to monitored production for enterprise agentic AI
  2. Part 1 CRISP-AG: An Artifact-Centered Framework for Enterprise Agentic AI Governance v3.0 An implementation framework of eight producible artifacts and a nine-phase lifecycle for governing enterprise agentic AI
  3. Part 2 Agentic PRD Standard v3.10.2 A hub-and-spoke requirements and evidence standard for enterprise agentic systems
  4. Part 3 Specification-Driven Design for Agentic Systems v1.0.3 From theory to practice: a method, a fully specified worked example and a build playbook for governed multi-agent AI systems
  5. Part 4 Enterprise Agentic AI Harness Specification v1.3 Twelve mandatory out-of-model controls and five lifecycle gates for enterprise AI agents, across the major agent stacks
  6. Part 5 Agentic Security Specification v1.0 Runtime and record security controls for enterprise agentic AI, and the evidence each must produce before it is called demonstrated
  7. Part 6 Vendor Control Specification v1.0 Nine controls for the vendor services that enterprise agentic AI runs on

Projects

Papers & case studies

Applied to PACCA. The general standard is Part 2 of the series →
Applied to PACCA. The general method is Part 3 of the series →

Patents