Braedn Heney

I have never been able to leave a process alone.

Not in a disciplined way. It’s closer to an itch. If I’m working inside something and I can see a better version of it, I can’t get on with the actual work until the better version exists. That’s most of what I’ve done since I got here, and it’s the only thing I’d claim as a method.

For the record, I was late on my first day. I was heads-down on my computer, phone face down next to me, and when I finally picked it up there was a Google Meet notification sitting there from twenty minutes earlier. I hadn’t known I was supposed to be on a call at all. I was on my way out the door to get food. I ran back inside and joined a meeting that had two people left in it. That was my manager’s first impression of me. I was eighteen.

Jen is my cousin, and there was no posting and no search. I don’t actually know what went into the decision to hire me, and I’m not going to guess at it. What I can tell you is that I was eighteen, I’d been building WordPress sites for a few years, and I’d been running my own marketing work. I’m also not going to pretend that door opened for me the way it opens for most people. What I care about is what I did once I was through it.

Why I See It This Way

I didn’t come to this through code. I came to it through wanting to know why.

Why people do what they do got me into psychology. Psychology got me into marketing. Marketing got me into UX and UI. That’s what got me into building. So I’ve never been able to look at a task as only a task. There’s a person and a business on the other end of it, and there is always something upstream producing it. Once you’re looking upstream, you can’t stop.

That’s the whole thing. Everything below is just what it looks like in practice.

The Sentence I Keep Using

I was hired to write content. The AI was already in use here and the prompts and rules were already written, so none of that started with me. What I did that first stretch was read what came out, find what was wrong with it, and rewrite the instructions until it came out right.

Which is where I got the sentence I’ve used ever since. If the output is bad, don’t fix the output. Fix what’s producing it.

Then Blake started, doing content transfer, and we were on calls constantly talking about what we were each working on. One day he walked me through what content transfer actually involved. Open an existing page, copy the design, build the next one by hand, every time, for every article. Nobody asked me to look at it. It just bothered me. So I built it the way I thought it should work, one template and a custom post type, on a single site, and then I showed people. You put the content in and the page exists. Build it once and every post after that populates itself.

Nobody fought me on it. I’ve found that’s usually how it goes. It’s much harder to argue with a working thing than with a suggestion.

In Over My Head

Chris was the first person to put AI in front of me, and one of the first things I did with it was rebuild the Nearby Now integration so it would work in Webflow, replicating their PHP plugin against our own web services backend.

That one was genuinely hard. It was the first time I’d deployed real code anywhere other than a WordPress site, and it was in TypeScript, which I’d barely used. All of it was new to me at once and it took me a while to get right. I mention it because I think that’s the actual price of admission here. Not talent. A willingness to work in the open on something you don’t know yet, for as long as it takes to know it.

This Isn’t a Task, It’s a Workflow

I spent a weekend being curious about MCP. There was no problem I was trying to solve. I’d been working in Cursor, read about it, and wanted to see what it could actually do.

Not long after, a one-off landed that needed doing. A batch of support tickets to move out of Zoho Desk and into Asana, a dozen manual steps each. Reply to the client, close the ticket, check whether a developer had already touched it and route it back to the same person, attach the right account manager. I had the time, so I helped.

And looking at it, the thought was: this isn’t a task, it’s a workflow, and workflows are things you can hand to a machine. So that became the first MCP server I built that did real work, with the right tools and the exact procedure written out for it to follow. You gave it a ticket ID and the whole thing happened in about a minute. Then the next one.

The bigger one started as a sentence to Parker. He’d designed something with AI and was about to hand it off to a Webflow team to build. I said, what if we don’t. What if we build the infrastructure around it and ship it as an actual site.

That conversation is the ground floor of Hydra.

The Part That Isn’t About Systems

For a stretch I ran our offshore team. A full workday, then standups at 10:45 at night. I go to bed early, so I set an alarm for 10:43 to keep from sleeping through it. That alarm still goes off. Sunday through Thursday, every week. I haven’t run those standups in a long time and my phone still buzzes at 10:43.

That stretch taught me the one thing I’ve learned here that has nothing to do with efficiency. You cannot fix people the way you fix a process. You can only be the thing you want them to be. Be kind and be firm. And when you don’t know something, say so out loud, share your screen, and work it out in front of everyone, because pretending costs more than not knowing ever did.

Roles, Not People

We went from hundreds of us down to around forty. Here’s how I actually see that.

We don’t lose people. We lose roles. People change roles. My whole approach has been to make things as efficient as they can possibly be, even when that means some job full of manual work stops needing to exist. That isn’t a threat to anyone willing to move. The people we lose are the ones who get attached to one way of working, and around here the way of working has never held still long enough for that to be safe.

Future Me

I’m twenty-one. I was hired here at eighteen to write words, and I now build infrastructure that a fair amount of this company runs on. That isn’t a sentence I could have predicted, and I don’t make plans far enough out to have predicted it.

What I do instead, when there’s a decision in front of me, is ask whether future me is going to be better off, happier, or proud that I made it. It’s a good filter. It kills most of the things that feel good right now and cost you later.

It’s the same question I’ve been asking about every process I’ve touched here. Not is this working. Things can work and still be wrong. Just: is there a better version of this, and am I going to be the one who builds it.