Case file 04Things I've built
SpaciaOS
Autonomous sales speed for luxury real estate: every enquiry called, qualified and booked.
Open in the office →The same case file, inside my interactive office
In shortI founded, designed and built SpaciaOS, an AI sales system for real estate that calls every new property enquiry within seconds, qualifies the buyer and books the viewing. The landing page and waitlist are up, and the app launches in November.
- SSituation
- Property buyers enquire when they’re excited, and a slow reply loses them to another agent.
- TTask
- SpaciaOS is my own product. I founded it and designed and built it: the product and dashboard, the landing page and waitlist, the brand mark and five teaser films, working with AI coding agents.
- AAction
- The landing page shows the speed instead of claiming it, agents can listen in and take over any call with one tap, the app shows only real data, and the waitlist opened before the app.
- RResult
- Designed and built the whole product with AI agents: dashboard, landing page, waitlist, brand mark and five teaser films. The waitlist is live ahead of the November launch, and the app runs on real data only, so the first numbers it reports will be real ones.
- Client
- SpaciaOS (my own product)
- Category
- Product · my own
- Role
- Founder: designed and built it (product, dashboard, landing page, brand mark, launch films)
- Team
- Me, directing AI coding agents
- Timeline
- 9 Sep – 8 Oct 2026 for the build (200 commits); launch planned for November 2026
- Platform
- Web app (sales dashboard) and a public landing page with a waitlist
- Tools
- Next.js, NestJS, Neon Postgres, Clerk, Resend, Cloudflare Workers
- Status
- Landing page and waitlist up; app launching in November
Property buyers enquire when they’re excited. Wait an hour and the excitement fades. Wait a day and they’ve called someone else. SpaciaOS answers every new enquiry from portals, ads, website forms and WhatsApp with a real phone call within seconds, day or night. It asks about budget, timing and financing, books the inspection into the right agent’s calendar, and writes up every call. Agents can listen in and take over any call with one tap.
SStage 1Situation
TStage 2Task: my part
01My part
SpaciaOS is my own product. I designed and built it: the product and dashboard, the landing page and waitlist, the brand mark and five teaser films. I worked with AI coding agents, and all 200 commits in the repo are mine.
AStage 3Action: the decisions
02Decision 1 · Show the speed instead of claiming it
The whole product promise is “seconds”, so the page proves it with motion. A running “CALLING 00:02 → CONNECTED ✓” readout, a lead card that moves from enquiry to call, and the qualification checks and agent calendar filling in. The problem section does the opposite: a pile of unanswered notifications and a timer that keeps counting.
03Decision 2 · The agents stay in charge
An AI that phones your buyers is a trust problem for an agency. So the product never replaces agents, and the page says so. Every call can be heard live and taken over with one tap, and every call comes with a transcript and a short summary.
04Decision 3 · Real data only
I set a strict rule for the app. It shows no mock or sample data: services start empty and show only what the backend returns. If the voice provider rejects a call, the user sees a plain-English error, and the call is never faked. The waitlist count is the real number of sign-ups, with no padding.
05Decision 4 · Waitlist before the app
Only the public pages are deployed: the landing page, the waitlist, a privacy page in English and French, and a 404. The dashboard and sign-in stay off the public site until launch. Prices on the landing page follow the visitor’s country (naira, dollars, pounds or euros), and five short teaser films end on the same line: “See you in November.”
RStage 4Result
- 1Founder: designed and built the whole product
- 5Teaser films for the launch
- Nov 2026App launch, with the waitlist already open
06Result so far
One founder designed and built the whole product with AI agents: the dashboard, the landing page and waitlist, the brand mark and five teaser films. The public pages are live (landing page, waitlist, a privacy page in English and French, and a 404), and the waitlist opened before the app so demand shows up first. The app launches in November and runs on real data only, so the first numbers it reports will be real ones.
Before launch: a voice number that can dial Nigeria, a custom domain, a legal review of the privacy page, and the full app and backend deployed.