Our recent cases. Just a sample, from startups to F500.
14 Tools to Automate Design
AI for animation:
• Runway Gen–4 – fast text-to-video with great motion realism.
• Kling – longer shots, smoother camera work.
• Wan – crisp detail and steady character movement.
• Veo 3 – cinematic control and consistent lighting.
• Sora 2 – extended clips with richer storytelling.
• Midjourney Video – quick style tests and mood explorations.
Figma plagins for different purposes
AI and generation:
• UX PILOT – generates user flows. Our team lives in it.
Assets and brands:
• Iconify – 275k+ icons with clean SVG import.
• Brandfetch – pulls logos, colours and fonts by company domain.
Content and data:
• Content Reel – ready strings, avatars, icons and your own libraries from Microsoft.
• Google Sheets Sync – maps copy and images from Sheets to layers by tags.
• CopyDoc – export or import copy to DOCX, XLSX, CSV and localisation tables.
Linting and cleanup for big design systems:
• Design Lint – finds inconsistent styles before handoff.
• Clean Document – removes hidden layers, flattens single groups, snaps to pixel.
UX Healthcare Europe 2025 Lecture
Really happy I got to speak at UX Healthcare Europe 2025! Shared a case how we used our experience of working on Tinder app to boost conversions for the telehealth app with 2M+ users :)
Thanks for having me!
Top 3 UX Mistakes in AI Products (2025)
It’s 2025 – the models keep getting smarter, but the user experience often hasn’t caught up. At Humbleteam we see the same problems again and again when working with AI teams.
1. Invisible feedback Users don’t know what the AI is doing or thinking. The black-box effect kills trust and drives people away.
2. Endless inputs Forcing users to write long, open prompts with no hints or shortcuts. Most freeze or drop off. If your AI doesn’t guide users with suggestions along the way, you lose them.
3. Over-promising, under-delivering Big promises on the landing page – then a confusing, generic, or disappointing first experience. Sometimes a user tries two or three prompts, gets poor results, and never comes back. (If you’re a PM, filter your analytics for users who only made 2–3 prompts before dropping off – and review what those prompts were. It’s eye-opening.)
None of these problems are fixed by better models. They’re fixed by better product teams.
We've Replaced Half of Our 3D and Motion Work with AI
What used to take days – or at least hours – in motion and 3D now takes minutes. The sketches you see attached were done in roughly 10-15 minutes, half an hour max.
What’s even more interesting is how the AI learns over time.
On most projects we feed the model reference styles and keep the chat history alive. The longer we stay on a project, the better and faster it gets at matching our visual language.
Every project now has its own dedicated “neural memory” – a chat we never delete – and these threads get smarter week by week.
Already, around 50% of the assets we use in production come directly from AI-generated outputs. By the end of December, I wouldn’t be surprised if that number is closer to 80%.
It’s wild to see creativity move this fast.
Why 500 Designers Shocked Me This Week
Yesterday reminded me how wide the gap is between designers who use new tools – and those who don’t.
We ran our regular UI Boost program. Two days, nearly 500 designers signed up.
My job was to show them how to make interfaces that are not only clean, but also engaging and quick to produce.
We asked every participant: what slows you down the most?
Over 60% admitted they barely use AI or Figma plugins.
For me, tools like ChatGPT, NanoBanana, Midjourney and plugins now take up half my design process. I spend less than 4 hours in Figma each day. The contrast is massive.
We put it to the test. Starting with a blank artboard, I mocked up a landing page in 17 minutes using AI and plugins. Students without these tools needed hours(!!!) for the same task.
After two days side by side, one thing stood out: some designers can create 10–15 website blocks in a single day, while others spend the entire working day on one single version.
That speed gap is staggering.
How Amazon and Apple Once Shared the Same UX
I’ve been preparing for a UX conference on conversions and came across one of my favourite stories – maybe the conversion story of all conversion stories.
Back in the late 90s, Amazon had the same problem most online shops still face today: too many abandoned carts and low conversion. Nearly 70% of carts were left behind, and average purchase conversion sat below 2%.
The reason? Checkout friction. Four or five steps to buy. Manual data entry. Far too many chances to change your mind.
Amazon’s answer was simple but radical at the time – 1-Click checkout.
Store customer details after the first purchase Collapse checkout from 5 steps to 1 Add trust signals – confirmations, instant feedback. The results were huge.
Conversion jumped from around 2% to nearly 10% – a 500% lift.
Cart abandonment fell by more than 40%. Average order value was +5%!!!
Apple even licensed it for iTunes and the App Store, paying Amazon for the privilege. And once the patent expired in 2017, every platform copied it.
It’s a great reminder – sometimes a small UX change can be worth billions.
The Difference Between Good And Great Designers
We work with a lot of design teams building sports apps, and there’s one mistake I see again and again. It looks small, but it quietly ruins great ideas.
Many designers, when a product owner asks them to design a feature, just design that feature - and stop there. That’s the mistake.
A great designer should always think about what happens before and after the feature.
Let’s take an example. You’re designing a football app and adding a prediction game before a match. Most designers stop there. A great designer asks:
• What happens five or ten days before the match? Could we send an email or push to let fans know the prediction game is coming?
• What happens during the game? If a fan’s prediction is wrong, can they change it at halftime?
• What about the off-season? Could we run quizzes or “prediction training” to keep users engaged?
Great design means thinking beyond the task. It’s about how the experience lives before, during, and after the interaction. That’s where great UX actually starts.
Day At Red Bull HQ
I had the chance to speak at the Red Bull HQ in California. It was such a privilege to be invited - and honestly, the vibe there was amazing.
About 12 people joined the session, where we shared what we’ve learned about fan engagement in 2025 after hundreds of conversations with sports fans and product teams.
But the most surprising part wasn’t even the talk itself - it was the place.
Right in the middle of the campus, there’s a massive fitness center where actual athletes train.
It’s surreal - you’re sitting there designing a new app or an Apple Vision Pro concept, and just behind the wall, someone is literally training for their next world record. That mix of creativity and raw sport energy feels special.
And of course, it’s pure California - sunshine, sport, tech, and great people. Love this mix!