Index of work

WORKIN THEOPEN

Things I've shipped where the interesting part was the constraint, not the framework. Most of it is source-available.

01Projects
01In build · 2026

Bubbo

A virtual pet for your sourdough starter.

Feed it and it visibly rises and domes; neglect it and it slumps and spreads, exactly as a real culture does. The sky follows the real time of day and the weather follows the starter’s mood — all derived from state and the clock, with no weather API and no location permission. Guided bakes start their own timers. Nothing leaves the phone.

Platform
iOS · Android
Built with
Expo · React Native
Your data
Never leaves the phone
Network calls
None
Open case
02Live · 2025

Webflow AI Image Enhancer

Enhance and generate images without leaving the Webflow Designer.

A Designer extension that puts Google's Gemini image models directly into the canvas — select an asset, enhance it, apply it. No round-trip through another tool.

Platform
Webflow Designer
Built with
Gemini image models
API key
Yours, not ours
Source
Open
Open caseSource ↗
03In build · 2025—

Agent tooling

The layer the studio builds its own products on.

Agent architecture, model routing, and interfaces that make an assistant feel like a colleague rather than a chat box. The console further down this page runs on an early slice of it.

Platform
Web
Built with
Next.js · LLM APIs
Models
Swappable
Status
Early
04Live · 2026

This site

A résumé you scroll through a camera.

Scroll-driven WebGL depth field, a noise-displaced horizon that reacts to scroll velocity, and an agent that has read the whole CV. Built to prove the front-end claims rather than list them.

Platform
Web
Built with
three.js · GLSL · Motion
First load
188 KB
Reduced motion
Fully supported
Source ↗

Morein the oven

Current work is AI automation and agent tooling under OLAI.ONE — mostly unshippable screenshots for now.

Start a conversation