Purpose

Improve services without surrendering control.

Public-purpose organisations should be able to improve how work and information flow without losing control of their data, decisions or future to unaffordable or opaque technology.

Our mission

Build open, trustworthy and demonstrable building blocks that help charities and public services:

  1. discover how fragmented work and information actually flow;
  2. connect systems through explicit, replaceable interfaces;
  3. automate bounded operational workflows with accountable human control; and
  4. retain evidence showing what happened, why it happened and how confidently the result should be treated.

Our ambition

Test whether a small team can apply enterprise discipline without enterprise bloat: clear ownership, event contracts, zero-trust boundaries, externalised policy, privacy by design, observable outcomes and repeatable assurance.

Success is not the number of components built. It is whether a demonstrator makes a meaningful organisational problem easier to understand, test and improve—and whether an informed outsider can inspect the evidence and reach their own conclusion.

Founding principles

How we hold the work to account.

01

Public purpose before technology

Start with the service problem, the people affected and the outcome sought.

02

Open by default, private by necessity

Publish reusable code and learning without exposing personal, confidential or security-sensitive material.

03

People retain accountable control

Assistance never silently acquires authority over consequential decisions.

04

Evidence over assertion

Separate ideas, demonstrations, measurements and production-ready claims.

05

Components own their decisions

Services communicate through explicit commands, events and interfaces.

06

Privacy and security are behaviour

Purpose, minimisation, consent, retention and auditability belong in the design.

07

Use technology deliberately

Select Rust, cREXX, Kubernetes and other tools for evidenced value, not identity.

08

Build the smallest decisive experiment

Create the minimum complete path needed to answer a meaningful question.