re-masked my living room with the Quest3, transforming it into a old-school Game Boy aesthetic.

Is this a world changing idea? Maybe not, but building and conceptualizing it in under a week is incredible. Just a bit ago, something like this might have taken months or even years for someone new to hardware development. The future is here 👀

Real-time augmented reality visualizing what a user might hear “walk straight through the phone booth to enter.” While we built this as an AI pendent to move away from glasses, these interactions could eventually extend onto AR glasses too.

The video feed runs through Gemini as well, for context-aware feedback, with wayfinding responses voiced by.@elevenlabsio. A companion web app, built entirely with @boltdotnew streams the camera feed to mobile and provides an on-demand UI when needed.

Here’s a closer breakdown on how the wearable works + tools used:. We used an ESP32 camera module to stream video, keeping the user localized through @NianticSpatial VPS for hands-free and accurate wayfinding. The physical device itself was prototyped using a Bambu 3D printer.

Built my first hardware + software project in just a couple days for our internal hackathon: AI wayfinding companion. Used Midjourney, Bolt, Supabase, Gemini, Eleven Labs, and Niantic Spatial VPS to bring it to life. The right tools make building insanely fast. How it was built:

RT @XRarchitect: Why not use the real-world as a game level?. This is a real-time augmented reality prototype using a physics based charact….

👉 Super simple web camera kit optimized for mobile devices. Perfect for testing out some AI vision ideas outside 👀.

Monday Prototype: A Gemini-powered workout analyzer. Built using my open-source Web Camera Kit boilerplate.

Testing. Midjourney -> Dither Boy

This is what three.js looks like in real life

This is the first time I’ve seen AI-generated video that actually feels somewhat usable for branding. Friday exploration with Midjourney.

You can try it yourself by putting in front of the GitHub URL. 👉 It’s a very simple kit at its core—and that’s the point. I got tired of hacking together web-based cameras and dealing with all the weird quirks (especially on

👉 Let’s take a closer look at some things I’ve recently built with it:. 1. AI workout coach – record a clip and get real-time feedback. Organize workout media in gallery. 2. Roast/Toast – snap a pic and Gemini roasts or hypes you up. 3. Travel guide –

One of my submissions for the @boltdotnew hackathon: an open-source web camera kit. I kept rebuilding similar components for some AI/computer vision apps. So I finally made a mobile-first boilerplate worth sharing. This + Gemini API = powerful, real-world AI vision apps.

Throne claimed. Midjourney reigns

Here is a splat of the above video 👀

I think I’ve fallen in love with Midjourney again. This is how I imagine world building.

I need to do more tactile stuff. This was a fun build.

Developed for the entrance of the @NianticSpatial booth! Come check it out. And built using @the8thwall Studio.

Web based augmented reality portal using VPS. Entire scene is 10mb in size. I learned a lot about baked lightning building this one. Hit me up if you are at AWE 👀