Skip to content

Process

Six steps, and
launch is only the fourth

Every project runs the same way. You always know what is happening now, what comes next, and what you get at the end of each step.

Step by step

What happens in
each step

  1. 01

    Discover

    Week 1–2

    We interview the people who will use the system every day, not only the people commissioning it, and audit whatever is running today. The goal is to find the real constraint before anyone opens a design tool.

    • Stakeholder interviews
    • Current-system audit
    • Scope document
  2. 02

    Design

    Week 2–6

    Research, information architecture and interface design, with the system architecture planned in the same weeks rather than after it. A screen the data model cannot serve gets caught here, while changing it is still free.

    • Flows and wireframes
    • High-fidelity UI
    • System architecture
  3. 03

    Develop

    Week 5–14

    Sprints with a review at the end of each one, so the first time you see the system is not the week it is due. Accessibility and performance are checked automatically on every commit rather than audited at the end.

    • Working software each sprint
    • API and schema
    • Test coverage
  4. 04

    Deploy

    Week 12–16

    Production launch with a zero-downtime strategy, and with backups and monitoring provisioned as part of go-live. Where an old system is being replaced, the two run side by side until the new one has earned the switch.

    • Production environment
    • CI/CD pipeline
    • Backup and rollback plan
  5. 05

    Manage

    Ongoing

    The subscription phase. Monitoring, security patching and bug fixes against a tiered SLA, through a support channel that reaches someone who knows your system rather than a queue.

    • SLA response times
    • Monthly security patching
    • Quarterly health report
  6. 06

    Scale

    Ongoing

    Feature and infrastructure expansion as the business grows: a second branch, a new department, a traffic profile nobody forecast. All of it planned against the same architecture, so growth stays a configuration change.

    • Capacity planning
    • Feature request pipeline
    • Infrastructure expansion

Principles

Four rules
on every project

Projects rarely fail at the end. They fail at the start, when nobody writes down what done means.

Agree what done means first

Before any design, we write down what the system must do on day one and how you will know it works. That list decides the scope and the quote.

A written “not doing” list

Every project leaves something out. We decide what in week one and write it down, rather than discovering it at launch.

Something to see every week

Rough work, shown weekly on your phone or a call. Problems surface while they cost a day, not a month.

Steps you can stop after

Each step ends with something you can use on its own. Stop at any point and you keep everything made so far.

Staying in touch

We work
where you already are

No portal to log into and no status reports to read. Questions go on WhatsApp, reviews happen on a call.

  • WhatsApp

    Day-to-day questions and quick checks

  • Google Meet

    The weekly review, recorded if you like

  • Figma

    Live designs you can comment on

  • GitHub

    The code, in an account you own

Ready when you are

Start with a free scoping call

Tell us what you need on a call or on WhatsApp. You get a written plan and a fixed quote, whether or not you go ahead with us.

Or email us directly at hello@uxatom.com