Process

From a sentence to a shipped product.

No discovery theatre, no ninety-page strategy deck, no surprise invoice. Six stages, each with a concrete output you can point at.

  1. Scope, honestly

    Days 1–3

    You describe the outcome you want. We come back with what actually has to be built to get there, what it will cost, how long it will take, and — importantly — which parts of your idea we think are wrong. If we are not the right people for it, we say so at this stage rather than three weeks in.

    • Written scope and assumptions
    • Fixed price or clear rate structure
    • Realistic timeline

  2. Model and architecture first

    Week 1

    Before any screen gets designed, we settle the data model and the architecture. What are the entities, how do they relate, what has to stay fast at scale, where does AI genuinely help, and what is the deployment shape. Getting this wrong is the single most expensive mistake in software, and it is nearly always made in week one.

    • Database schema
    • Architecture decisions, written down
    • Repo created in your GitHub

  3. Ship something real, early

    Week 1–2

    You get a working URL as fast as possible — deployed, real, and usable, even if incomplete. Feedback on a live thing is worth an order of magnitude more than feedback on a mockup, and shipping early forces the infrastructure questions to get answered while they are still cheap.

    • Live staging URL
    • Continuous deploy from your repo
    • Working core flow

  4. Build out, in public

    Weeks 2–8

    Iteration against the live version. Every push deploys, so you can watch it take shape rather than waiting for a reveal. We work in visible increments and you can raise a problem the day it appears instead of at handoff.

    • Continuous deploys you can watch
    • Regular written updates
    • Open issue tracker

  5. SEO, performance, and legal

    Pre-launch

    Structured data across every page type, generated sitemaps, canonical URLs, Open Graph coverage, and a real Core Web Vitals pass. Plus the parts everyone forgets: privacy policy, terms, cookie disclosure, and an accessibility pass against WCAG 2.1 AA.

    • Full schema and sitemap coverage
    • Performance budget met
    • Legal pages in place

  6. Launch and hand over

    Launch week

    DNS cutover, SSL, analytics, monitoring, and a handover document that explains how the thing actually works. Everything lives in your accounts — your GitHub, your Railway, your Cloudflare — so the handover is a permissions change, not a migration.

    • Production launch
    • Handover documentation
    • Full ownership transfer

Ownership

You own all of it, from day one.

This is not a differentiator we invented for marketing copy — it is the only arrangement we think is defensible. Agencies that hold your code hostage are solving their retention problem with your business risk.

  • Source code Pushed to a repository in your own GitHub account, from the first commit.
  • Hosting Deployed in your Railway workspace, billed to you, under your control.
  • Domain & DNS Registered and managed in your Cloudflare account, with your credentials.
  • Data Your database, your records, exportable in a standard format at any time.
  • No lock-in Standard open tooling throughout. Nothing proprietary to migrate off.

Ready to start at stage one?

Stage one costs you nothing and takes about three days. You come away with a written scope and a real number, whether or not you decide to go ahead.

Get a scope See the work first