A year ago I moved to Dubai with my wife and our son, who's now two and a half. It hasn't been a quiet year. The conflict in the region made things unpredictable for a while, and we've just moved house again. But we're settled now. The new place has a separate level for my office, which changes everything when you live with a toddler. Waking up, looking out the window and seeing a Rolls-Royce, a Bentley, a Lamborghini and a few Porsches go past, then hearing the familiar rumble of a G-Wagon, is a decent hit of motivation too. What I keep noticing here is the people. Everyone I meet is positive, driven and quick to offer help. There's an energy I didn't feel back in Australia. Running my own studio from here is the setup I always wanted. I can go hard on the work and still be there for my son.

Such a timely announcement. Seconds before myself and others jumped on the trend of switching from Framer to custom. This is what I needed.

Tried directing Claude Code into our Figma design system three months ago. Unusable. Shelved it. Tried again recently on a live client project. It's executing better than I expected. The gap between "doesn't work" and "works" is now about a quarter. Plan accordingly.

Last year I shared the spin the chair test. Push back from the desk, turn around, spin back and notice what your eye lands on first. I still do it on every screen we ship. These days it's mostly on products founders have built with AI tools. They tend to fail it the same way. Everything's the same weight. Four buttons all asking to be clicked. Nothing tells you where to look. So I spin, see what I see, and fix that one thing first. Usually it's picking the action that matters and letting the rest step back. Try it on whatever's open right now. What did you see first?

It's actually crazy trying to nail down the perfect offering right now. The way we work is changing, which means our processes need to adapt, and the operational/pricing models we've had (especially for the studio) need to change. I'm going to just try embrace the constant fluidity and hope to attract clients that are also open to the changing times. We can work it out together.

How many designers and studio founders are in Dubai? I run Halaska Studio from here. A year in, I've only occasionally bumped into great people, but I'm realising there are a lot more of you around. Reply or tag someone. Would love to meet more of you.

A founder showed me five brands his team built with AI. "Good enough to be out there." Then he pointed at the fonts on each one and said even fixing those would change everything. He's right. AI gets you to 80%. The 20% is taste, and taste is what people pay for now.

Exploring

Directing Claude Code into production Figma only works because two things were already true. The design system was built and perfected by humans first. And I can articulate direction after 20 years of briefing designers. Take away either one and it produces junk. Keep both and it's shipping.

After 20 years in design, six of them at Google, I've spent the last year going headfirst into AI at Halaska Studio. It's in how we design, how we run the studio, what we ship, and how we build MVPs for clients. And it's changed what I think the future of agencies and designers looks like, now that AI is getting properly good at design. AMA.

One of our latest clients booked one discovery call and we started the following week on a $20K a month retainer. No pitch deck. No proposal round. They'd seen our work and knew my background, and they wanted us hands-on straight away. We rebuilt their design system and component library from scratch, then moved on to their core flows, fixing the ones that mattered and designing new ones. Two syncs a week with their team kept us on track and the feedback flowing, which is why we could move as fast as we did. The best clients already know what they want. They just need the right people to do it.

In January I said design studios would look completely different by the end of the year. Nine months on, here's where mine landed. We built our own system for how the studio works: 20 years of product design foundations with the client's brand on top. We now set it up inside each client's team, and it's become as important as the work itself, because it's what keeps their output on-brand after the engagement rolls over. Some studios I referred to back then have closed. Some are flying. The ones flying all changed what they sell, not what they charge. Called it. Now building it.

A founder paid us $50K a month last year when they could have used AI and a couple of freelancers. I asked why. "We're not paying for design skills. We're paying for perspective that challenges our assumptions." Six years at Google and 20 years of shipping taught me that the tools were never the edge. Everyone has the tools. The calls you make with them are the edge. That founder raised $9M off the deck we built together, and the product is now live. It's the whole bet behind Halaska 2.0.

The temptation to rebrand everything to The Design Office of ...

One of the tricks I picked up from working at Google is a very simple way to make sure your interface feels on-brand: Start with the core brand colour. Let's say it's blue. You create a palette of blue shades from light to dark, almost white to almost black, all with a blue tint. The text isn't just pure white. It's a very light blue. The background isn't pure black. It's a very dark blue. Using these throughout the core elements will create an almost unnoticeable but present brand 'feeling' through the interface. It works especially well for trading interfaces where the positive/negative colours can easily dominate, making it hard to incorporate your brand without it being jarring. This technique works perfectly for any product and industry. And if you're building with AI, you can just prompt it in: "Take my brand's base colour and generate a tinted scale from near-white to near-black. Use those tints for every background, surface, border and text colour in the interface instead of pure greys."

I started Halaska Studio in Dec 2023. It grew faster than I expected and taught me more than I expected. If I was starting again today, here's what I'd think harder about. Studio or freelancer. I built a studio because that's what I thought you did. But most clients came to work with me. A clearer solo route would have removed a lot of overhead. What one person can do now. With AI in the process, a single senior designer covers what took a small team two years ago. The maths on hiring has changed. Pick the lane early. Crypto founders were my first clients and my best referrals. I widened too fast and diluted that. I'd still start it. I'd just start it smaller. What's shifted for you?

One of the best client relationships we've had at the studio wrapped up this year. A year and a half working directly with the founder and their internal team, fully embedded, across product, brand and website. We shaped the product strategy, built the brand from the ground up, and designed the core product end to end. Over that time it grew into a mid six-figure engagement, which says more about the trust we built than anything I could. I got close to that team. Still proud of what we made together. They're in stealth, so I can't show the product yet. When it launches, I'll share what we built.

Mentoring is pretty much at capacity right now. A slot or two opens up when someone reschedules, so it's worth checking back. Paid coaching (45 min, $299) still has room: chrishalaska.com

A client once flew me to Tuscany for a strategy offsite. A week in a villa, surrounded by vineyards. Fresh pasta for lunch. Planning the product with their team. We worked together for a year and a half through the studio, and that week is still the one I think about. You get more done in a few days around one table than in a month of calls.

Take it from me. The culture of the workplace is the number one thing when it comes to enjoying your work. This post is an indication of the toxicity.