Vibe Coding Didn't Solve the Hard Part

or, We're back -- sorry it took so long.

It’s been over a year since I wrote anything about Helen’s Foundry. In that time I graduated with my MBA, bought a cat cafe, remodeled two bathrooms, and failed to launch my startup.

That last one is the reason for the silence.

Building software is harder than I expected. Especially when it’s two people with other obligations and no salaries and the finish line keeps moving. But honestly, the harder part was the shame. We raised a modest amount from friends and family — not enough to pay us, but enough to cover development and infrastructure — and I couldn’t face the people I’d taken money from. Saying “whoops, I failed” felt impossible, so I just... didn’t.

Instead I polished. Redesigned the same flows, the same screens, over and over. Technically better every iteration (more usable, more visually coherent, whatever metric you want) but not launching. Not putting a product in front of anyone. A really cozy way to waste a lot of time feeling productive while accomplishing.. absolutely nothing.

I know I’m not alone in that. The illusion of productivity is seductive as hell. I made design improvements! I pushed PRs! I read startup articles! And I was no closer to solving a real problem for a real person.

In mid-April I decided to go back and look at what we’d actually built. Not to resurrect it, but to confirm it was dead. “AI ate our lunch” is an easy thing to say, and it’s partly true, but it was also a cop-out. I failed to lead. I failed to execute my own vision, implement my own strategy. That’s on me.

But re-reading the artifacts we’d spent months building, something shifted. The AI boom hadn’t made Forge irrelevant, it had made the problem clearer. People who know how to build things still often don’t know why. Vibe coding collapsed the time between shower thought and working prototype. It never solved the foundational problem: writing things down and ruthlessly interrogating every assumption.

So I started over. From scratch, by myself. I built a “Forge Lite” in three weeks and liked where it landed. I dropped a link in Slack and tagged my cofounder (who I’d essentially ghosted for months) to ask what he thought. He saw it too. Then he did what Justin does: within a month he’d taken my yolo proof of concept and turned it into something that actually deserved to launch.

We’re opening early access today. No charge, no trial, no subscription. Just your honest feedback. Send an email to hello@helensfoundry.com or request access at Helen’s Foundry.