The process
What working with us actually looks like — day by day.
No fake-it-til-you-make-it agency theatre. No 47-page SOW. Here’s every step you and we take, in order, with what ships at each.
Discovery
01
Day 0 · free
Drop your .bubble export
AI reads every workflow, names every risk, and ships you a McKinsey-grade rebuild plan with real numbers. ~10 minutes end-to-end. No call, no signup.
What ships
- 25-exhibit report (web + print)
- 12-slide PPTX deck for your CTO
- Live preview deployed to a Vercel URL
- tailwind.config.js seeded from your palette
Your only job
Click upload. That's it.
Discovery
02
Day 1-3 · free
30-minute scoping call
We walk through the report together — what to preserve, what to rebuild, what to throw away. You ask anything; we answer with the report open on screen. End of call, you get a fixed-price proposal in writing.
What ships
- Written proposal with milestones
- Fixed price per phase
Your only job
Book a slot. Show up. Skeptical questions welcome.
Kickoff
03
Day 4 · paid
Repo, accounts, sandbox
Within 24h of signed proposal we hand you a GitHub org with three repos seeded (app, infra, runbook), a staging Vercel project, a Supabase sandbox, and a Slack/Linear channel — whichever you prefer.
What ships
- GitHub org transferred to you
- Staging Vercel + Supabase wired
- Daily standup at a fixed time
Your only job
Tell us your preferred standup time + chat tool.
Build
04
Week 1
First slice — auth, billing, schema
We start with what your users would notice missing first: signup, login, password reset, Stripe checkout, the user/account/subscription tables. By end of week, this is on staging and you can use it like a real product.
What ships
- Working auth on staging
- Stripe sandbox subscriptions live
- Postgres schema migrated 1:1 from your Bubble data model
- Integration tests covering the critical paths
Your only job
Hit the staging URL daily. Tell us what feels wrong.
Build
05
Week 2-3
Spine workflows
The 5-10 workflows that ARE your product — order placement, message send, content publishing, whatever moves the business. Each ships as its own PR with its own demo video.
What ships
- Each spine workflow on staging with a demo loom
- Cron jobs and background queues replicated (Inngest, not Bubble scheduling)
- First slice of your data dual-written so we can cut over without downtime
Your only job
Approve each PR. Reject what's wrong.
Build
06
Week 4-N
Long tail + polish
The remaining 70-150 workflows — most of which are variations of a pattern, not net-new. We batch them by similarity (all the 'list-with-filter' pages together, all the 'edit-record' forms together) so each batch ships in days, not weeks.
What ships
- Every page from Bubble rebuilt or explicitly deprecated
- Performance regression tests on critical user flows
- Accessibility + lighthouse scoreboard
Your only job
Pick what to deprecate, not just what to rebuild.
Migrate
07
Final week
Cutover
We've been dual-writing data for weeks, so cutover is a DNS change + a 30-minute window. Bubble app stays running in read-only mode for 30 days as a fallback. You're the one who flips the switch.
What ships
- DNS pointed to Vercel
- Bubble app frozen in read-only
- 30-day rollback plan documented
Your only job
Click the deploy button. We watch the dashboards together on a call.
Aftercare
08
Day +1 to +30
Hand-off + warranty
We're on call 30 days after cutover for any production issue at no extra cost. Daily morning Slack check-in, on-call rotation in case something pages at 3am. After 30 days, you keep everything; we're an email away if you want us back.
What ships
- 30-day on-call coverage
- Runbook for every critical path
- Daily standup until you say stop
Your only job
Ping us when something looks off. We answer the same day.
Start at Day 0. It’s free.
We’re telling you exactly what happens at each step because we’ve done it dozens of times. Drop the file or grab a slot — whichever’s cheaper for you.