§ masterplan.md
AI Room Redesigner
Upload a photo of any room and get unconventional redesign ideas with shoppable furniture tags and color palettes.
01
App overview and objectives
Upload a photo of a room, choose a style, and get photoreal redesign options with every visible item linked to something you can actually buy. The objective is to close the gap between inspiration and purchase in a single session.
02
The problem worth solving
People know their room feels wrong but cannot describe the fix, and Pinterest boards never map to products available locally. Turning a photo into a shoppable plan removes both the imagination gap and the sourcing work.
03
Target audience
- Primary — First-time home owners and renters furnishing on a budget.
- Secondary — Independent interior designers who want a fast client-facing mock-up tool.
04
Roles and permissions
- Visitor — read-only public pages, can sign up.
- Member — owns their own records, cannot see other users' data.
05
Core features
- Room photo upload with automatic room-type and lighting detection
- Style picker (minimal, warm, industrial, traditional) with intensity slider
- Side-by-side before/after with a drag divider
- Detected item list with matched shoppable products and price totals
- Save designs to a project board and share a link with family
- Budget filter that re-generates a cheaper version of the same look
Deliberately later
- 3D walkthrough of the generated room
- Local carpenter and installer marketplace
- AR preview on phone
06
Technical stack
React front end, server functions for generation, object storage for photos, a Postgres catalogue of products.
Why: Image generation is slow and expensive, so it belongs in a queued server job with progress feedback; storage and catalogue are ordinary database work.
Alternatives: Pure third-party redesign API (fastest to launch, no control over product matching) versus training your own model (months of work, only worth it at scale).
07
Conceptual data model
User — account, saved projects, budget preference
RoomPhoto — original image, detected room type, dimensions estimate
Design — source photo, style, generated image, cost estimate, status
Product — name, price, retailer, category, image, affiliate link
DesignItem — design, product, position hint, alternative options
08
Integrations
- AI image generation
- File storage for uploads
- Retailer or affiliate product feeds
- Payments if commissions are collected
09
UI design principles
- Break the grid on purpose: overlapping layers, oversized type, one surprising interaction per screen.
- Motion carries meaning — elements should morph between states instead of hard-cutting.
- Keep exactly one experimental idea per page so the interface stays usable, not chaotic.
10
Security considerations
- Room photos are private by default; sharing creates a separate token-based link.
- Strip EXIF location data from every upload before storage.
- Signed, expiring URLs for all user images.
11
Development phases
Phase 1 — Prove the core
- — Room photo upload with automatic room-type and lighting detection
- — Style picker (minimal, warm, industrial, traditional) with intensity slider
- — Side-by-side before/after with a drag divider
- — Static content and design system in place
- — Basic analytics
Phase 2 — Make it real
- — Detected item list with matched shoppable products and price totals
- — Save designs to a project board and share a link with family
- — Budget filter that re-generates a cheaper version of the same look
- — Accounts, sign-in and password reset
- — Empty, loading and error states everywhere
Phase 3 — Polish and launch
- — Performance, accessibility and SEO pass
- — Legal pages, contact route and 404 handling
- — Wire up: AI image generation
- — Wire up: File storage for uploads
Phase 4 — Grow
- — 3D walkthrough of the generated room
- — Local carpenter and installer marketplace
- — AR preview on phone
12
Challenges and solutions
Risk — Generated rooms look impressive but unbuildable
Solution — Constrain generation with the original geometry and only ship designs whose items exist in the catalogue.
Risk — Product matching accuracy
Solution — Start with a curated catalogue of a few hundred items rather than the whole internet.
Risk — Generation cost per user
Solution — Free tier of three designs, then credits; cache by photo hash and style.
13
Future expansion
- Whole-apartment planning
- Designer marketplace with paid consultations
- White-label version for furniture retailers
14
Page list (13 pages)
- 01 PUBLIC Upload — Photo intake, style choice and generation progress.
- 02 PUBLIC Gallery — Saved and public designs, filterable by style and budget.
- 03 PUBLIC Shop — The shoppable item list behind any design.
- 04 PUBLIC Designers — Directory of humans who can finish the job.
- 05 PUBLIC Pricing — Credits, free tier and pro plan.
- 06 AUTH Sign up — Create an account with email or a social provider.
- 07 AUTH Log in — Return to the account, with error and lockout states.
- 08 AUTH Reset password — Request a reset link and set a new password.
- 09 APP (signed in) Account settings — Profile, email, password, language and delete account.
- 10 PUBLIC Contact — Contact form plus real address, phone and email.
- 11 LEGAL & SYSTEM Privacy policy — What data is collected, why, and how to remove it.
- 12 LEGAL & SYSTEM Terms of service — Rules of use, liability and account termination.
- 13 LEGAL & SYSTEM 404 not found — Friendly dead end with search and links back.
15
Page map
PUBLIC AUTH APP (signed in) LEGAL & SYSTEM ──────────── ───────────────── ─────────────────── ─────────────────── ├─ Upload ├─ Sign up └─ Account settings ├─ Privacy policy ├─ Gallery ├─ Log in ├─ Terms of service ├─ Shop └─ Reset password └─ 404 not found ├─ Designers ├─ Pricing └─ Contact key flows: Upload ──▶ Gallery Gallery ──▶ Shop Shop ──▶ Designers Upload ──▶ Pricing
