Skip to content
webappsdev
Menu

Process

A clear process, from first conversation to launch and beyond

Five steps, each with something concrete for you to see and approve. You always know what happens next, what it costs, and what you will get.

  1. Step 01

    Discover

    We talk about your business, your customers and how the work gets done today: the spreadsheets, the email threads, the steps that take too long. You get a short written plan covering what to build first, what it will cost to run each month, and what can wait.

  2. Step 02

    Design

    Using my AI-assisted design workflow, I turn the plan into screens you can click through and a data model you can review. Changing a design costs minutes, while changing finished software costs days, so this is where we get it right.

  3. Step 03

    Build

    Development happens in short iterations on a private preview site, so you see progress as it happens and can change course early, while changes are still cheap. The AI workflow speeds up the routine work, and I review and test every change.

  4. Step 04

    Launch

    The app goes live on Google Cloud with HTTPS, backups, monitoring, analytics and search-engine setup in place from the first day, not added as an afterthought months later.

  5. Step 05

    Support & grow

    Your team gets documentation and a walkthrough. After that I can keep maintaining and improving the app as the business grows, or hand it over cleanly to your own people. The code is written so that either option works.

The in-house AI workflow#

I started using AI in software development early, and over time I have built it into a workflow of my own. It is not a chatbot writing code unsupervised. It is a set of tools and proven patterns that I direct and check at every step.

What the AI does#

  • Turns requirements into designs. Notes from our discovery conversation become screen layouts, a data model and an API outline, ready to review in days rather than weeks.
  • Generates code from proven patterns. New code builds on structures I have already used in production, with the same validation, error handling, security headers and logging in every project.
  • Writes the tests alongside the code, so every feature arrives with checks that it works and keeps working.

What I do#

  • Decide the architecture and the trade-offs.
  • Read and review every change before it is merged.
  • Run security scans and test the result the way your users will use it.
  • Keep your customers’ data out of the loop. The AI works on specifications, designs and code.

What it means for you#

  • Shorter timelines. Routine work happens at machine speed.
  • Lower cost. Less time spent typing boilerplate means more of your budget goes into the features that matter to your business.
  • Consistent quality. Every project gets the same tested, secure foundations from the start.

What you will need to provide#

  • Someone who knows the process the app will support, available for a short conversation each week or so
  • Access to the tools to integrate with, such as a Google Workspace admin for sign-in or shared Sheets
  • Quick feedback on prototypes, which is where your input saves the most money

Got a spreadsheet that has outgrown itself?

Or an idea for a customer portal, or an old app that needs rescuing? Tell me about it. The first conversation is about your business, not about technology.

Start a project