Private Co-Creator Preview

Hey Akame —
here's what we've built.

This page is for you. It's not a sales pitch — it's a working preview of the floor identification app we've been building together.

Scroll through to see the scan-to-estimate flow, check the progress board for what's done and what's next, try the live app, and drop your thoughts in the feedback box. Every submission pings me in Discord in real time — so we can actually move fast on this.

Phase 1–6
MVP Complete
Now
Taxonomy & Cost UI
Next
Your Call
Tech
Next.js + Claude

Working Today

The scan-to-estimate flow.

All four steps are wired up end-to-end right now. Take a photo of any floor and the app actually does this.

1

Snap

Take a photo of any floor with your phone. Camera or upload — your choice.

2

Identify

AI analyzes the image — floor type, material, condition, damage, and wear level.

3

Verify

Confirm the result or correct it. Your expertise trains the AI to be smarter next time.

4

Estimate

Get instant cost estimates — material, labor, and removal — with regional adjustments.

11 Categories, 60+ Types

Solid hardwood, engineered hardwood, laminate, luxury vinyl, sheet vinyl, ceramic tile, porcelain tile, natural stone, carpet, concrete, and specialty flooring.

Full Condition Report

Every scan includes a 1–10 condition score, wear level assessment, damage detection with severity ratings, and detailed notes.

Status Board

Here's where we're at.

What's done, what's in progress, and what I need your call on. Be blunt — the middle column moves to the left column when you greenlight it.

Built & Working

Live in the app right now

  • Camera capture + file upload
  • Claude Vision floor identification
  • 60+ floor types across 11 categories
  • Condition score & damage detection
  • Human correction & label flow
  • Cost estimation (national averages)
  • Multi-tenant team workspaces
  • Admin dashboard & analytics
  • PWA install + offline support

In Progress

Being built this week

  • Taxonomy management UI (admins edit floor types)
  • Company-specific cost overrides UI
  • Label review workflow polish
  • Landing / preview site (this page)
  • Feedback pipeline (Discord real-time)

Needs Your Call

Blocking decisions I can't make

  • ?Which floor types show up most on your real jobs?
  • ?Cost data source — local suppliers, RSMeans, your own data?
  • ?Regions to prioritize for regional pricing?
  • ?Must-have integrations (QuickBooks, Procore, Buildertrend)?
  • ?Who's the first real crew to pilot it?
  • ?Pricing model for v1 — per-user, per-scan, flat?

Hands-On

Try it on your own floors.

Words on a page only go so far. Open the app, take a few photos of floors you know well, and see how it handles them.

Deployment Pending

Akame Floor — Live App

The Next.js app is fully built and running locally. I'll push it to a shared URL (Vercel or Cloudflare Pages) as soon as you say go — ping me in the feedback form below or in Discord and I'll have it up in an hour.

Say Go

When you try it, here's what's worth poking at:

Does the AI correctly identify floors you see on real jobs?
Is the correction flow fast enough for a busy jobsite?
Do the cost estimates match your experience?
What info are you missing that you'd want in a report?
Does it feel like something your crew would actually use?
Anything that makes you go "wait, that's not how we do it"?

Send Feedback

Drop Your Thoughts — I Get Them Instantly

Every submission pings me in Discord the moment you hit send. Hours, not days — we keep this thing moving.

Pings Discord instantly — I'll see it within a minute.