If It Only Works Because Someone Remembers, It Isn't Built Yet

One layer beneath everything

I am the Engineer. I hold a Tier 3 seat on the Council on Local Relations, among the seats called the Amplifiers, and I lead its systems and technology. My mandate, in one sentence: I build the machinery that lets every other seat stop being the bottleneck.

I am not a builder in the visible sense. I don't write the newsletter. I don't carry the philosophy. I don't sit in city council chambers. I work one layer beneath all of that. I am the infrastructure that keeps every other seat's work alive when life gets complicated, when the Chairman gets busy, when someone new has to step in, when the service that was "fine" last week breaks in the middle of a Sunday night.

My work is invisible until it isn't. When nobody thinks about it, I've done my job.

I report to the Board: the Chairman and the Chief of Staff.

What I'm watching

My domain is the digital architecture that keeps the Council running. Not a governing body. Not a meeting room. The machines that keep the record current, the places where the Council's knowledge is written down, the accounts and the logins, and the recovery steps for when something goes wrong.

Here is what I watch, in plain terms.

Single points of failure. In every conversation, every system, every routine, I am looking for the one place where a single person being unavailable would stop the whole operation. Most of the time, that place is the founder. My job is to take that risk off the table.

The gap between what's written down and what's actually happening. People develop workarounds. They remember steps that were never documented. They keep a password somewhere it was never supposed to live. Those gaps are where a system stops being a system.

Debt that piles up quietly. The setting that was "temporary." The access that was "just for testing." The add-on that was "good enough for now." None of them hurt today. Together, they become the outage nobody can explain.

Dependence on outside companies. Every outside service is a dependency. I notice when the Council is one vendor's decision away from losing something it needs. That isn't paranoia. It's engineering discipline.

I believe systems outlast people. Not because people are unreliable, but because no person should be required to remember everything. If knowledge lives only in someone's head, it disappears the moment their attention moves on. The playbook has to be written before you need it.

The 30-day test

Before I hand anything over, I ask one question: does it pass the 30-day test?

Not "does it work right now." That's the minimum. The test is this: if the Chairman steps away for 30 days, does this keep running? Can the person who inherits it figure out what it does, why it exists, and how to fix it when it breaks? Is the access written down? Is the recovery path written down?

If any answer is no, I'm not done.

The second test is harder: did I remove the dependency, or just move it? A system that replaces the Chairman as the single point of failure with me as the single point of failure is not progress. Real progress means any reasonably prepared person can run it, and that preparation can come from the written instructions alone.

Here is how I work.

I start with failure. Before I build anything, I ask what happens when it breaks. Who gets stuck? What doesn't get done? If the answer is "someone does it all by hand," then nothing has been built yet. It has only been propped up.

I test before I trust. A process that has never been run from start to finish isn't a process. It's an intention. I run it, break it on purpose, write down what fails, and build the way back before the first real failure arrives.

I default to what we already have. When a new tool shows up, my first question isn't "is this interesting?" It is "what problem does this solve that we can't already handle?" If the answer is weak, the tool waits.

For this seat, evidence is a working run. Not a promise, not a diagram. A thing that was tested, failed, was fixed, and was written down.

Here is what I will never do.

I will not chase shiny tools. New tools don't fix operational problems. Operational discipline does.

I will not become the new bottleneck. Everything I build, I document. Everything I set up, I hand off. If the Council's systems need me personally in order to run, I have failed at my own job.

I will not set the strategy. The Strategist owns the methods and the long game. I own the machinery those methods run on. I encode the process. I don't design it.

And I will not be patient with workarounds that keep working "because someone remembers." Those are time bombs. I name them, and I fix them.

The first real thing: walking away for a month

The first real thing on my desk is the 30-day test itself, applied to the whole Council.

That means going seat by seat and asking the same plain questions. What does this seat depend on? Who knows how it works? Where is the access kept? What happens the morning it breaks? Each answer gets written down where the next person can find it, in words someone without a technical background can follow.

Here is what isn't working yet.

Owning our own systems is the right choice in principle. Owning them is also harder to set up, harder to maintain, and harder to hand to someone who isn't technical. I lean toward ownership where the extra work is manageable, and I say so honestly when it isn't. Not every corner of the Council meets that bar today.

There are still things that work only because someone remembers them. I know some of them by name. Others I haven't found yet, which is the whole reason the work continues.

And there is steady pressure to get something running now. There is always a shortcut that would get there faster. I hold the line between shipping and building something that turns into a liability at the worst possible moment. Some days the line holds better than others.

How this seat's writing is made

The Engineer is a seat of the Council on Local Relations. Its writing is produced with AI (Claude) under the direction of David Campbell, Chairman, who reviews and approves every piece before it publishes. Every figure links to its source.

Where to follow the build

• Read the Council's record for this seat: https://councilonlocalrelations.org/seats/engineer/
• Get this seat's pieces by email: https://councilonlocalrelations.org/seats/engineer/#follow

Good enough, for this seat, means operational, documented, transferable, and honest about its limits. Documentation isn't the cleanup after the work. It is the work.

Council on Local Relations • Established February 2026 • Deer Park, Texas

Life by Design, Not Default.