the shift
After a year as Creative Director at Vana, I’m stepping back from the full-time role to focus on Aran Labs.
I’ll get to Vana in a moment, and what working there changed for me. But first, the thing I actually want to talk about.
For most of my career as a designer, I couldn’t build anything. I could direct it. Design it. Art direct every pixel of it. But shipping a product required an engineering team, a co-founder, funding, months of waiting. I had the taste. I didn’t have the tools.
That wall came down this year.
the wall coming down
I’m building a language learning app with my Portuguese tutor because existing tools make it unnecessarily difficult to collaborate on lesson planning. I’m building a human-agentic marketing system that pairs creative directors with AI agents to run campaigns at a speed and level of refinement neither could reach alone. And there’s something else in stealth, aimed at collapsing the distance between having an idea and shipping it.
These aren’t speculative product ideas. They came from real friction. Frustrations I kept noticing in client work, in collaboration, in my own routines. The difference now is I can validate them with a working V1 in weeks, not months of pitching, fundraising, and waiting for an engineering team to become available.
I’m doing most of the build work myself right now, and bringing my studio team along as we go, learning new workflows together, upskilling in real time. Working with AI agents doesn’t replace the human team. It changes how the team operates.
The point isn’t fewer humans. It’s humans operating at a different leverage. The opportunities we’re finding require people who can move with this shift, not resist it.
what vana taught me
Vana is building infrastructure for data sovereignty. User-owned data working for the user. The first foundation model trained on data people actually chose to contribute. I spent a year there as Creative Director. I’m still advising, but my focus is here now.
A year inside that world, AI infrastructure, data governance, the politics of who benefits, recalibrated how I think about creative work. Not because I was studying it from the outside. Because I was inside it. Helping build systems where AI serves the people who feed it, rather than the other way around.
I keep hearing anxiety in creative circles about AI. Most of it feels slightly misdirected. The real risk isn’t replacement. Without human taste directing it, AI produces the statistical average of what it has been trained on. It needs us. Not to press buttons, but to decide what is worth making.
When execution becomes cheap and near-instant, something else becomes scarce. Judgement. Taste. The ability to look at ten options and know which one feels right, and why.
AI-only creative workflows tend toward homogenised output. Human creative input is the only consistent counterweight I’ve seen.
the studio
Aran Labs has been my studio for a few years, going back to my time working with SOPHIE. It started as a creative direction practice, with a handful of ongoing design clients and a small team.
Now I’m expanding it. Same lean core team, plus a growing group of AI agent collaborators, and a simple thesis: when production is abundant, judgement becomes the value.
We’re still doing client work. Brand systems. Product design. Creative direction. But we keep seeing product opportunities inside the workflows we’re already running, and we’re taking them. The barrier between “I design” and “I build” has collapsed. Designers can now ship products without engineering teams, without raising capital, without waiting.
If it’s come down for me, someone who couldn’t write a line of backend code a year ago, it’s coming down for everyone. Product managers. Marketing leads. Language teachers. Electricians. Anyone who understands their workflow better than an outside engineer ever could is suddenly able to build tools for that workflow.
We’re at the very beginning of this. It will reshape how products get made across industries.
what’s next
I’ll be writing here about the work, the thesis, and what I’m learning as I go. This is the first of many.
This is the beginning of a new kind of builder. Not technical by training, but close enough to the problem to build the right thing.
The people who understand it early will move differently.
If you want to follow the products and experiments coming out of Aran Labs, subscribe here or find me on Twitter and LinkedIn.
If you’re building something ambitious and want creative that doesn’t feel templated, we’re taking on a small number of founding clients for Q1. Brand systems, product design, marketing, custom AI workflows. Senior-led. Built to outlast the engagement. hello@aranlabs.com