There are lots of ways to talk about building technical depth.
You can run training sessions. You can swap tool recommendations. You can talk about AI workflows in theory.
Or you can clear the calendar for a day, ask everyone to make something weird, and make sure it ships.
That’s what the Algolia Product Design team did in a recent hack day built around a simple brief.
“Create something fun, slightly ridiculous, and genuinely usable for someone else on the team. Use AI tools however you like. By the end of the day, it has to run on the web.”
No Figma-only outcomes. No polished prototype as the final deliverable. The result had to be live, clickable, and shareable.
That one requirement changed the exercise completely.
Instead of stopping at ideas, everyone had to work through the full flow. Concept, prompting, code generation, editing, debugging, deployment, and demo. The point wasn’t to turn designers into engineers in a day. The point was to build real confidence with end-to-end AI tools by making something small and shipping it.
The setup
The format was intentionally lightweight.
Each participant filled out a short personal brief with prompts like:
“I waste time on…”
“My current obsessions are…”
“Describe yourself using 3 emojis…”
“A theme I love is…”
“What annoys me daily is…”
Then names were drawn, so each designer built for a specific teammate.
We also added a “ridiculous constraint” for each project. One designer had to make something that looked like it came from 1996. Others had to include multiple sound effects, use only one color, avoid text entirely, or make the experience get worse the more you use it.
That gave everyone a useful mix of structure and freedom. Nobody had to invent a prompt from scratch, but nobody was boxed into making something sensible either.
Why this worked
The official goal was simple, everyone ships a small app, game, or website using AI coding tools such as Claude Code or Cursor, then deploys it somewhere public.
But the more interesting goal was confidence.
AI tools change who gets to build, how quickly ideas become interfaces, and how far someone can get before they need a handoff. For designers, that makes experimentation much more immediate. Instead of describing an interaction, you can test it. Instead of stopping at a prototype, you can put it in a browser and see what happens.
That kind of hands-on learning matters. Algolia’s developer persona research shows that technical users value customizability, quick implementation, and AI support that helps them build real experiences. A one-day hack day turned those ideas into practice.
It also helped that nobody was building for a generic user. They were building for James, or Kim, or Shar.
That made the work more personal, more playful, and more thoughtful. People weren’t aiming for broadly impressive. They were aiming for something that would feel unmistakably right for one person on the team.
What the team built
The best outcome is also the simplest one: everyone shipped.
And the projects were every bit as funny, specific, and charming as you’d hope.
Verdict is a dramatic little decision machine for purchases, ideas, and questionable life choices. You enter a prompt and get scored on deeply unserious but extremely satisfying dimensions like substance, style, and delusion. The whole thing has a glossy sci-fi confidence that makes it feel like a future computer taking itself far too seriously. It’s sharp, funny, and just self-aware enough.
Bonus constraint: must include multiple sound effects
Gastonagotchi is a Tamagotchi-style game starring Gaston, a very expressive virtual french bulldog with big Monday energy. You help him stay balanced as he deals with naps, moods, and general life management. It’s playful, noisy in the best way, and full of the kind of overcommitted charm that makes a small browser game instantly lovable.
Bonus constraint: it gets worse the more you use it
This one has one of the strongest titles of the day and fully commits to the bit. It frames dinner planning like a survival challenge from a ruined future, with a tone somewhere between post-apocalyptic terminal and chaotic meal generator. It takes an ordinary question and gives it maximum drama, which is exactly what a good hack day project should do.
Kim’s Happy World is a cheerful memory game built around little details from Kim’s interests. Players flip cards, match pairs, and uncover references to things she loves, from food to travel to music. The emoji-only rule makes it even more playful. It feels personal without being sentimental, and silly without losing the care that went into it.
Be Trail turns hiking recommendations into paranormal case files. Trails become mysteries. Outdoor planning becomes a lightly haunted investigation. It’s a great example of what happens when a practical idea gets pushed through a strong point of view. You still get the useful details, but now they arrive with cryptid energy.
Weather.exe takes a familiar utility and gives it unapologetic retro desktop-software energy. The concept lands immediately. It’s weather, but through the lens of old-school interface nostalgia, with all the visual commitment that implies. It’s simple, recognizable, and exactly the kind of idea that gets better once someone decides to really lean into the aesthetic.
What stood out most
The strongest moment wasn’t just the demos. It was seeing people react to what had been made for them.
Each project landed in roughly the same way: half laughter, half recognition.
There’s something powerful about making for a specific person. You stop aiming for “good” in the abstract. You aim for something that will make them smile. That changes the work. It makes it more observant. More generous. More memorable.
It also makes experimentation feel safer. The work doesn’t have to be roadmap-ready to be successful. It just has to be thoughtful, finished, and real.
Why we’d do it again
Hack days can easily become vague or overproduced. This one worked because it had just enough structure:
A clear deliverable.
A hard deadline.
A live deployment requirement.
A personal brief.
A ridiculous constraint.
That combination kept the day focused without making it feel formal.
More importantly, it gave the team a practical way to learn by doing. Designers didn’t need to master every tool. They just needed enough time, a reason to try, and a format that pushed them all the way to shipping.
That’s often where the real learning starts.
What made the day work wasn’t just the AI tooling. It was the combination of pressure, play, and specificity. People had a reason to make something, someone real to make it for, and just enough structure to get out of their own way.
The result was a set of strange, thoughtful, fully shipped projects, and a team that now feels more confident turning ideas into working things.
Other Types
%20(2).svg)
