Skip to content

About Jeep Marshall

150

Jeep Marshall LTC, U.S. Army (Retired) Airborne Infantry | Special Operations | Process Improvement 📧 admin@herding-cats.ai


I served 26 years in the U.S. Army, spanning airborne infantry and special operations. Seven of those years went to training brigade-level staffs through simulation-driven exercises — where every planning framework described in this series got tested under operational pressure, repeatedly, with real consequences for failure.

I hold Lean Six Sigma Black Belt certification. I’ve applied DMAIC methodology to military operations, organizational systems, and — more recently — to AI agent architecture and workflow design.

Currently I’m building a comprehensive “second brain” knowledge management system using Obsidian, Claude AI, and multi-agent workflows. That system accidentally became the laboratory documented in Paper 3.


The AI industry has a discipline problem, not an intelligence problem.

I learned this in the Army. Individual brilliance doesn’t survive contact with organizational complexity. A platoon of exceptional soldiers fails without doctrine. A fleet of brilliant AI agents does the same thing — just faster and at greater expense.

These papers make the case that the next generation of AI practitioners needs Black Belts and battle staff officers more than another language model.


I didn’t plan to write seven research papers and a case study. I planned to build a better personal knowledge management system.

What happened instead: I started applying military planning frameworks to AI workflows, ran into every coordination problem the academic literature predicted, documented the failures, and kept refining the doctrine. By the time I had 1,768 git commits across 33 days of multi-agent operations, the papers were writing themselves.

Claude AI assisted throughout. The research, the analysis, the live field tests, and the cross-series synthesis all happened in real-time sessions. The ideas and the operational experience are mine. The execution partnership made them publishable.


Each paper applies a multi-disciplinary analytical team — a methodology I borrowed from how a proper battle staff operates:

  • LSS-BB (Lean Six Sigma Black Belt): Waste analysis, process capability assessment, DMAIC framework
  • QASA (Quality Assurance Standards Analyst): Source verification, cross-reference validation, publication readiness
  • ASS2 (Automation, Structure & Scalability, Safety & Security): Three-domain review applied together — automation potential, structural scalability, and safety/security posture. Flat, equal-weight framework, not single-lens analysis.
  • Creative Arts Practitioner: Hands-on field testing, visual evidence, professional workflow assessment
  • Observer-Controller: Scope management, quality gate enforcement, cross-series continuity

This isn’t decoration. The multi-perspective lens produces findings that a single analyst misses.


The series targets both practitioner and academic audiences:

  • Small Wars Journal (military AI doctrine, operational applications)
  • Military Review (command and control frameworks)
  • War on the Rocks (strategic analysis)
  • Journal of Force Structure (organizational design)
  • General practitioner audience via Obsidian Publish

For questions, collaboration, or speaking inquiries:


← Back to Home | Glossary & Acronyms →


© 2026 Jeep Marshall. All rights reserved. Herding Cats in the AI Age” is an original research series by Jeep Marshall.