~/prototyping
Fully working prototypes within 2–3 days
Validate your idea and your business model before you commit a budget to it. Real software, real feedback, real fast.
01 / how it works
From idea to evidence in three days
Day 0 — Sharpen the idea
A focused call to cut your idea down to the one thing worth validating. We define what “working” means before anything is built.
Day 1–2 — Build
AI-first development with a senior engineer’s judgment: a fully working prototype, not a clickable mockup.
Day 3 — Validate
You put it in front of real users or stakeholders. We iterate on what the feedback actually says — or you walk away having spent days, not months.
02 / why it works
Prototypes beat pitch decks
A prototype answers the questions a slide deck can only speculate about: will users get it, will they use it, will anyone pay for it. An AI-first approach makes this affordable — twenty years of engineering experience make it reliable.
And because it’s real software, a successful prototype is a head start, not a throwaway: the learnings (and often the code) carry directly into the actual product.
03 / proof
Recent examples
high-five-habits ↗
A habit-building Lovable prototype I use daily with a few friends — built Lovable-only, with a fully working Supabase backend behind it.
email-archive ↗
Instant keyword + semantic search over twenty years of my own mail, in one Docker container that never phones home. Python, SQLite, Vue, and an MCP endpoint for AI assistants. Open source.
Have an idea worth testing?
Tell me about it. If it can be validated in three days, we’ll do exactly that.