Skip to content
Book a call (opens in a new tab)

How we work

We build it with you, then we make it last.

One deliberate process, run by one team. We design the architecture and tech stack first, so we can prototype fast on foundations that hold. Then we iterate the working system through to production, and stay on to run and grow it.

Architecture first, so we can prototype fast and support it for years.

  1. Blueprint 3 to 5 days, plan and price
  2. Architecture data, security, stack, hosting
  3. Prototype working system within weeks
  4. Production live review, then hardening
  5. Run and grow support, DevOps, product team

Prototype and production iterate in blocks, pause when you need to

One Growth Pilot team runs every step

Product leadership Architecture Engineering AWS-certified DevOps

The process

Five steps, in order. Each one makes the next one faster.

  1. Blueprint

    3 to 5 days

    We work with you on the problem, the users and the constraints, and leave you with a plan and a fixed price for the first block. You can stop here and keep the plan.

  2. Architecture and stack design

    Before any screens

    We design the data model, security, integrations and hosting, on a proven stack. This is the part that is expensive to get wrong and hard to fix later. Doing it first is what lets us move fast afterwards without rebuilding.

  3. Rapid prototype to working system

    Within weeks

    We build on those foundations inside your business, next to the people who will use it. Your team is using real software within weeks, not reading specs.

  4. Iterate to production, in blocks

    Block by block

    Each block: the thinnest complete path first, a two-week live review on real work, then production hardening. Security, backups, monitoring, performance and the edge cases that only show up under real use. Pause between blocks if you need to.

  5. Run and grow

    Ongoing

    Ongoing system support from AWS-certified DevOps, and a fractional product team that keeps your product evolving as your business grows.

Inside a block

The thinnest complete path first. Two weeks on real work. Then hardening.

  1. Weeks 1 to 3

    First build

    The thinnest complete path through the problem, on the architecture already in place.

  2. Two weeks, on real work

    Live review

    Your people run actual jobs through it. We watch, listen and log what matters.

  3. Weeks 6 to 8

    Second build and hardening

    Shaped by what the review showed. Security, backups, monitoring and performance made production-grade.

  4. End of block

    Sign-off

    Reviewed, in use, supported. Pause here or scope the next block.

The live review is what finds the expensive mistakes in week six instead of month six. What does not fit gets changed while there is still time and budget to change it.

Quality levels

We name the level up front, so the price and the plan match.

Working prototype

Real software your team uses on real work, built fast inside your business. Enough to prove the shape and the value.

Where every build starts

Production for your business

Hardened, monitored, backed up, security reviewed. Reliable enough to run the business on, with support after launch.

What most clients buy

Sellable to other companies

Multi-tenant, billable, documented and supportable as a product in its own right.

Quoted separately, later

How we build

A stack we know, and AI-assisted engineering with senior review.

A common, proven stack

Most of what we ship runs on the same building blocks: TypeScript, React and Next.js on the web, Expo / React Native for mobile, Postgres (usually via Supabase), and AWS for hosting, storage and document processing. Reusing a stack we know is what makes the architecture step fast and the systems easy to support.

AI-assisted engineering, with senior review

We build with Claude Code and Cursor. Every repo carries its own agent instructions, coding standards, a domain glossary and architecture decision records, so the AI tools follow the project's rules rather than guessing. Reusable agent skills cover common workflows such as migrations and new features. Architecture and every significant change are still reviewed by our senior engineers. Prototype speed without giving up production quality.

Tools we use

Claude Code Cursor Supabase AWS Expo Next.js

Worked example

Monk Earthworks: from a prototype in the office to a crew app in daily use.

A civil contractor running on WorkflowMax and spreadsheets. We designed the foundations, built the back office first, ran real claims through it with the office team, then shipped a timesheet app that 13 crew use every day. Iteration continues block by block.

Read the Monk Earthworks story

Start with a Blueprint.

Three to five days with Kevin. A build-ready plan and a fixed price for block one, with the fee credited to the build.

Book a 30-minute call (opens in a new tab)