Hospitable gave us one job. Make a short-term rental host feel seen and use story to drive a feeling.

No feature walkthroughs, just a video that a new short term rental host would relate to.

What we landed on was a voicemail. A host leaving a message for a friend who just listed their first place. And for the visuals we landed on approaching it from the POV the host's iPhone. What are moments a host would actually capture in the lead up to hosting for the first time?

We made the whole thing with generative AI. That's the part some people will push back on, and that's fair.

But most of our time went into restraint. Pulling the generated stuff back until it felt ordinary and lived-in instead of shiny. AI just let us build that world without a five-figure shoot getting between us and the idea. Multiple models for the homes and the people, Seedance for video, and VO with elevenlabs.

It was refreshing to use these tools to chase a feeling, not just try to wow people with AI. We learned a ton in this project and look forward to creating more with this team.

This is so cool congrats man

What if you could feel a photo?

Millions of blind and low-vision people can't - the images we love stay locked behind a screen. So we built Felt: drop in any image and it becomes a tactile relief with braille, ready to 3D-print. Felt redraws your picture into clean, raised lines, writes a braille caption and turns it into a real 3D file you can hold.

Felt was built in Figma. Claude shaped the idea how a flat image becomes a raised relief, and how to map each shape into readable braille - and through Figma MCP it worked right on the canvas with us. In Figma Design we built the whole product: the palette, the type, every component and state and the full flow from landing to printable file. We packaged it all into a Figma Make kit, so the design stayed a single, on-brand source of truth. Figma Weave generated our artwork as clean SVG and Figma Make turned everything into real, working code. Under the hood, Supabase with Nano Banana Pro redraws any uploaded image into clean, printable lines through a single AI API call. Then we took it into the third dimension: a 3D relief methodology turns those lines into a height-mapped model with real depth - rendered live in Three.js and exported as a 3D-printable file.

It's a side of Figma you rarely see - not just screens but 3D and design that leaves the screen for good.

Accessibility isn't a feature here, it's the whole point. Design doesn't have to stay on the screen. From screen to skin. 🤍

Try Felt: https://scene-border-57045515.figma.site

Figma Community Design: https://www.figma.com/community/file/1649037227670473474

Social post: https://x.com/kategate01/status/2067207561024041291

#ConfigMakeathon @figma

this is such an important tool! nice work!