The same capability, sold once more into the job next door
Open the real page ↗interiorai.com · alive as of 2026-08-31
Upload a photo of a room, get restyled versions back. Technically this sits on the same generation stack as the portrait product, but it aims at a different job — and that made it a separate product with separate revenue. The complaint that AI output all looks alike is answered here not with better technology but with a narrower target. ⚠️ Revenue figures are founder-published, so we do not print them (grade B).
In plain words
Send a photo of a room, get pictures of that room decorated. Same technology as the portrait product, different customer — so it sells separately.
The whole thing at a glance
To whom
People decorating or listing…
What
AI SaaS
Found how
X — the founder's public wor…
Paid how
Subscription and credit-base…
Flow: To whom, then What, then Found how, then Paid how
- Founder
- Pieter Levels
- Country
- NL
- Team
- —
- Startup cost
- —
- To first revenue
- —
- Automation
- high
- Reported revenue
- Revenue not disclosed
What is sold
Subscription and credit-based image generation
Pricing is published on the official site (liveness confirmed 2026-08-31). ⚠️ We do not copy the amounts across.
Who buys
People decorating or listing a home, plus property and interior professionals
What arrives is not what you keep
- What the customer paidThe number on the invoice
- AI toolsA share of the monthly subscription
- Platform feeDiffers by platform
- TaxDiffers by country and business form
- What is leftAt or below zero, it is not a price
This case does not publish its own tool costs, fees or tax. So we show the order and the items only — put your own numbers into the calculator.
AI's part and the person's part
Already had
Same founder as the previous case. The starting line already included a working product and its users. This is not a from-scratch condition.
If this is large, the AI is not the reason
What the AI did
Input is a room photo, output is a styled concept. Pick a style, the space is redrawn. The shared stack with the portrait product is the cost advantage — one capability learned, sold twice.
the part a tool replaced
What the person did
The human decision was choosing which job needs this picture, and when. Listings, preparing a rental, planning a move — moments right before a decision, where an image carries weight. Product separation came from the choice of application, not from the model.
stories that omit this do not replicate
How customers were found
X — the founder's public work
Search traffic from people looking for interior concepts
Cross-referral from the existing product's users
Paying customers
Pricing is published on the official sit…
Flow: X — the founder's public work, then Search traffic from people looking for interior concepts, then Cross-referral from the existing product's users, then Paying customers
How we verified this
Public dashboard or platform figure. May be self-reported
Confirmed — the service is live (official site HTTP 200, 2026-08-31) and the founder is named. Not confirmed — revenue. The quoted figure sits in the same founder blog post as the previous case, alongside numbers for other products. Pulling a single product's figure out of that mix is itself the risky step, so we publish none of it.
- Official site — liveness confirmed (HTTP 200) (official_site · checked 2026-08-31)
- Founder blog — figures for several products in one post. Supporting evidence only (founder_interview · checked 2026-08-31)
What made it work
Re-aiming one capability at an adjacent job is the cheapest second product you will ever build. New customer base, no new technology to learn. And it is the practical answer to sameness of AI output. The answer is not a better model; it is a narrower point of application.
Risks
- Concept images drift easily into pretty but unbuildable. Ignore real dimensions and structure and it stays entertainment, never reaching checkout.
- Used on property listings it can shade into misleading advertising.
- Revenue is reported mixed together with the founder's other products. Do not treat any figure as this product's own.
Does this work in Korea?
Among these ten, this transfers to Korea comparatively well. Rental turnover is high, listing photos visibly affect deals, and people want to see a concept before committing to a quote. ⚠️ The risk is letting a concept be read as a finished result. Label generated images as generated and state plainly that the built outcome may differ — in Korea that notice matters more than the render does.
Solo-feasible: yes · Automation: high
7-day action plan
Step 1. Collect ten local listing photos and note which ones actually make you stop scrolling
Step 2. Ask two working professionals when in their process a concept image would be useful
Step 3. Draft the disclosure line stating the image is generated, before you need it
30-day action plan
Step 1. Pick one space type only and tune output quality for that type
Step 2. Do ten free jobs for agents or fitters and collect what was missing each time
Step 3. Put the concept-versus-build disclosure in front of checkout, then open sales
Run the numbers
We do not print a single projected figure here. Put your own numbers in instead.
Get the next verification result
We check the AI income stories people are talking about, grade the evidence, and send you what we found — including what we could not confirm.
Words you may not know
- Verification grade
- How far we checked this case. A means the revenue was confirmed by someone other than the founder; B means the business exists but the revenue is only the founder's word.
- Revenue
- Everything that came in. Tools, fees and tax still have to come out before it is what you keep. The two are not the same.
- MRR
- Money that arrives every month, as opposed to a one-time payment.
- Self-reported
- A number the person stated about themselves. Nobody else confirmed it, which is why we do not print those numbers.
- Replicability
- How likely someone else is to get a similar result. The more the founder already knew the field, the harder it is to copy.