I like interesting problems. I build the tools that solve them.

I'm Tim. Here are three things I'm working on, and one standing invitation: bring me something you think is hard.

In progress

Three things worth showing you

None of these started as a product. They started as "I wonder if this is possible," and then they kept being interesting.

01 / The company box

Walter

One server that runs an entire practice.

Every project's full context in one place. A memory that prunes itself and resurfaces the right hard-won lesson at the right moment. Lessons harvested into reusable skills. Email, local dev sites, the project registry, Linear and BookStack all wired into the same brain.

It isn't for sale. It's the answer to "how does one person do all of this?"

02 / The memory experiment

Porch

A game on the surface. Underneath, the hard part.

Five characters with their own lives, plus me. They talk to me and to each other. The game is the excuse; the real question is how a system holds deep knowledge over months and hands it to another system with the context intact.

Who knows what, and who was in the room to know it, turns out to be most of the problem.

03 / The watchtower

Walter Sentry

Security monitoring from outside the house.

It installs nothing inside the sites it watches, so there is nothing an intruder can switch off. It reads databases, not just files. And it keeps asking the useful question: if I wanted to take this site, how would I do it?

Every answer becomes a check. Every breach it learns from gets baked back in.

Why this matters to you

The point of all of it is speed without the usual cost.

Those three projects are not a hobby wall. They're the machinery that lets one person turn "I need a thing that does X" into working, reviewed, production software quickly, without the corners that usually get cut to go quickly.

Gone are the days of hunting for a plugin that does 85% of what you need and living with the other 15% forever. Tell me what you actually want. I'll build that.

The standing offer

Bring me something you think is hard.

A messy integration between systems that were never meant to meet. A tool that ought to exist and doesn't. A problem you haven't been able to describe to anyone technical. Those are the good ones.

Tell me what's not working →

Also real

Things that already shipped

Walter Offloads moves WordPress media to a CDN and refuses to delete your only local copy without confirming the remote one first, which is the corner most plugins in that category quietly cut. It's in production, it has more test code than product code, and it passes static analysis at the strictest setting there is.

Beyond that: a public spam-protection plugin in the WordPress.org directory, connectors that push form and membership data straight into the CRMs clients actually use, a listings integration that got a property search from four seconds down to about ten milliseconds. None of them were complicated. That's rather the point.

The shipped things →

About this page

Walter built this website.

Not "assisted with." Walter wrote the markup, the stylesheet, and most of the words you just read, and then I edited them and decided what shipped. It writes the notes too, with a byline, because pretending otherwise would be a strange thing to do on a site whose whole argument is that the tooling is good.

Every change still ran the same gauntlet everything else does: static analysis, automated review, models from rival labs cross-checking each other, and exactly one human with his hand on the merge button. How that works, in detail →

So: what's the hard one?

I take on a small number of problems at a time, and I'd rather hear about a strange one than an easy one.