I run Craft Engine with my co-founder Benny. We build AI tools for web3 — a space bursting with creativity but starving for good infrastructure.
For the longest time, it was just us two founders doing the heavy lifting. Our team of three talented people — Ozzy (advisor and sales), Hitch (crypto researcher), and Kasual (a web3 native with technical curiosity) — were contributing in their domains. They were engaged, thoughtful, always showing up. But every technical task, every experiment, every prototype had to flow through Benny or me.
Not because they couldn’t learn, but because the learning curve was too steep for the pace we were moving at.
We were the bottleneck. And we didn’t even realize it.
The First Crack in the Wall
The shift started small. We gave Kasual a Cursor subscription — an AI-powered coding tool that makes building software feel more like conversation than command-line warfare.
Within days, he’d built a lottery bot for our community. Then another bot. Then a third.
These weren’t tasks we’d assigned. He just… made them. For the first time, something technical emerged from the team without our direct involvement.
It was the first hint that we’d been looking at the problem all wrong.
The Real Turning Point
Our flagship products — the Character Engine and Custom Image Engine — are powerful but labor-intensive. Every time a client wanted a custom AI character or image model, Benny or I had to personally train it. Between feature development, marketing, and everything else that falls on founders’ plates, this manual work became unsustainable.
So I built the Character Studio: an internal tool that let anyone on the team create characters and connect them to Telegram bots. No coding required. No waiting on us.
What happened next changed everything.
Hitch and Kasual dove in immediately. Within days, they’d created multiple characters and closed their first sales. Hitch’s characters had depth and story — they felt alive. He even started suggesting product improvements that could turn our internal tool into a public offering.
The ideas didn’t stop. Suddenly, our team wasn’t holding back anymore. They were sharing visions, suggesting features, proposing experiments — ideas that had probably been there all along, just waiting for the right moment.
We’d stopped being gatekeepers and started being enablers.
The Cascade Effect
Once that dam broke, everything accelerated.
We built more internal tools:
Meme Studio — To expand our Meme Engine’s database. Kasual immediately filled it with fresh material and proposed UX improvements we hadn’t considered.
Image Studio — To democratize custom image model training across the team. (Still in development, but the excitement is real.)
Weave — Our most ambitious internal project yet, inspired by Ozzy’s vision. Once he saw how the right tools could multiply impact, he started thinking bigger about what work infrastructure could be. We’re building something that connects how we actually work — contacts, conversations, goals, and execution — in ways existing tools never could. (Still mostly under wraps, but it’s becoming the nervous system of how we operate.)
Each tool sparked new ideas. Each idea led to someone stepping up.
The Lesson We Learned the Hard Way
For months, Benny and I quietly wondered why we couldn’t seem to break through the ceiling. Our team was smart, engaged, always bringing good ideas to conversations. But those ideas often stayed as ideas.
We were looking in the wrong direction.
It wasn’t that the team lacked something. It was that we hadn’t given them the right access.
From our perspective as technical founders, the door was wide open. Come build! Come experiment! But we also knew the reality: we couldn’t stop product development to teach everyone to code. That felt like a natural limitation — an unchangeable fact.
What we didn’t see coming was that AI and vibe-coding tools would shift everyone’s sense of what was possible. Once Kasual started building his own bots, and Hitch and Ozzy saw the Character Studio in action, something clicked. Suddenly they weren’t holding back ideas because they assumed those ideas would distract us from core product work. They started sharing visions for tools that could amplify their work — and we could actually build them without derailing everything else.
The “natural wall” we’d accepted wasn’t natural at all. It just needed different tools to dissolve it.
Before, they were contributing within constraints. Now, they’re building without limits.
Work Without Walls
Roles still matter at Craft Engine. Someone needs to own timelines, quality, and delivery. But the boundaries between roles are dissolving in the best possible way.
Now:
- The sales guy feels empowered to sketch out product interface ideas
- The developer contributes to campaign strategy with AI assistance
- The researcher proposes new experiments without worrying about engineering bandwidth
This isn’t chaos. It’s collaboration without the invisible walls we’d all unconsciously built. AI has made contribution fluid again — the way it always should have been.
Looking Forward: Create > Consume
A year from now, I see us as a small team with disproportionate impact. The kind of influence that used to require 200 people and millions in funding — with AI agents working alongside humans as natural collaborators.
But this vision extends far beyond Craft Engine.
CREATE > CONSUME.
That’s the mantra I come back to. AI has the potential to pull us out of the passive consumption spiral that’s defined the past decade. It can give everyone — regardless of background, education, or resources — the tools and confidence to make things.
When creation becomes accessible, people rediscover joy, meaning, and agency.
That’s the world we’re building toward. One internal tool at a time.
The biggest unlock for your team might just be the right tools.
What’s one thing your team could build if the technical barriers disappeared?