AI fails the way the environment lets it fail.
I'm Tabron Merryhill. I build governed AI systems using the principles civil engineering has relied on for a century — controls before execution, defined load limits, and human authority at every point a decision becomes irreversible.
An engineer's answer to a problem most people treat as policy.
My background is in supply chain systems, construction and civil engineering management, and applied AI. Across six years of deploying AI in real operations, I kept seeing the same thing: systems didn't fail because the model wasn't smart enough. They failed because no one built the environment the model was supposed to operate in safely.
So I built it. The result is the Merryhill Protocol and a portfolio of governed systems running in production — each one applying load-bearing engineering logic to software that makes consequential decisions. I work as Tabron Merryhill Systems Division, and I am honest about scale: this is focused, hands-on work, not a large firm. You deal directly with the person who designs and builds the system.
The Merryhill Protocol
A model-independent, platform-independent AI governance framework. It is built on five axioms, five architecture layers, and the L1–L4 irreversibility register shown above. Its central rule is simple and borrowed directly from engineering: governance exists before execution, not as an audit after something has already gone wrong.
Because the framework doesn't depend on any single AI model or vendor, the governance survives when the model changes. It has been deployed in production across construction, operations, and other domains — the same schema holding across each, which is the point.
Two environments. One principle.
Responsible development — of systems or of people — is never the result of isolated effort. It's the result of intentionally designed environments that make responsible growth possible. I build both. Design the environment first, because the quality of the environment determines the quality of what develops inside it.
The Governance Environment
The technical and organizational conditions that let AI operate safely and accountably — the infrastructure built inside an organization before AI is trusted with consequential work. Its purpose isn't to control AI; it's to create the conditions under which AI can act responsibly.
- Merryhill Protocol
- Enablement Platform
- Governance workflows
- Audit trails
- Human checkpoints
- Role-based approvals
- Deployment gates
- Continuous review
The Learning Environment
The developmental conditions that help leaders, operators, and organizations grow from ungoverned to governed thinking. Informed by how people actually develop new ways of thinking — so the people change before the organization does.
- AI That Works workshops
- Executive briefings
- Mapping Intensives
- Before Execution podcast
- Industry presentations
- Executive education
- Courses & certifications
Straight answers.
Who is Tabron Merryhill?
An AI governance architect and the founder of Tabron Merryhill Systems Division. I build governed AI systems using principles from civil engineering and construction management — controls before execution, defined load limits, and mandatory human authority over irreversible decisions. I hold an MBA and graduate certificates from the University of Missouri and an Applied AI credential from MIT Professional Education, and I created the Merryhill Protocol.
What is the Merryhill Protocol?
A model-independent, platform-independent AI governance framework organized around five axioms, five architecture layers, and an L1–L4 irreversibility framework that sets how much human authority each AI decision requires. Its core principle is that governance must exist before an AI system executes. It has run in production across construction, operations, and other domains.
What does Tabron Merryhill Systems Division do?
It builds governed AI systems and operational intelligence tools. The work includes the Merryhill Protocol governance framework, AI Search Discovery — which makes a business's content readable by AI search engines — and SitePulse, a daily check-in and reporting system for builders managing subcontractors.
Do you offer AI governance consulting?
Yes — governance gap assessments, AI opportunity mapping, control architecture design, and executive briefings for organizations deploying AI. The approach treats governance as an engineering discipline, not a policy document you file and forget.
Governed systems, in production.
Merryhill Protocol
The core governance architecture — intake, control review, irreversibility classification, human override, and a recursive audit trail. Deployed across multiple domains without rebuild.
AI Search Discovery
Makes your content readable by AI search, so when customers ask Google, ChatGPT, or Perplexity for what you offer, your business is the answer.
Visit AI Search Discovery →SitePulse
A daily check from every subcontractor, compiled into one morning report — so builders catch cost and schedule problems while they can still act on them.
Visit SitePulse →Governance Consulting
Governance gap assessments, AI opportunity mapping, and control architecture for organizations putting AI into real decisions. Engineering discipline applied to AI risk.
Start a conversation →Let's talk about what you're deploying.
Whether you're putting AI into consequential decisions or you want your business found by AI search, start here.