A team that owns the product, not a ticket queue
We take architecture, delivery, testing and the long-term evolution — working as your product engineering function rather than a supplier waiting for a specification.
Thirteen months in, the codebase is still absorbing significant new capability at a steady weekly cadence — which is the real test of whether the architecture was right.
Marine planning platform · still shippingWhat we build
Four surfaces. Most engagements need three of them.
Backend APIs and services
Layered .NET services with interfaces and dependency injection at every seam, so business rules stay testable and independent of transport and storage. Command and query paths separated where the load justifies it, not everywhere by reflex.
Mobile apps, offline-first where it matters
One codebase reaching iOS and Android. We have built for people working where there is no signal — dockside, inside nurseries, in the field — where the app has to hold the work and reconcile it later without losing anything.
Web platforms and internal tooling
Server-rendered React or Blazor, virtualised tables that stay responsive over very large datasets, shared filter systems, live progress over SignalR. The unglamorous details that decide whether a team actually adopts the thing.
The parts nobody demos
Schema migrations, CI/CD pipelines, health checks, structured logging, single sign-on and role-based access, automated test suites at unit, integration and API level. On one platform that means 550+ test cases, including a suite that asserts the business rules directly.
Four things that decide whether it lasts
None of these are visible in a demo. All of them are visible three years later.
The rules come from the people who hold them
On the marine planning platform, most of the operating logic existed only as practice. We ran it down case by case with the client’s specialists and wrote it into code with the reasoning documented inline — the rule file reads as a specification as much as an implementation.
One team across the whole stack
Because we own the database, the API and the UI, changes that would normally need three-way coordination between three suppliers are a single coherent piece of work.
Reviewed delivery, continuously
3,600+ commits over thirteen months on one platform, every one through a peer-reviewed pull request with integration tests as a gate — while the system stayed in daily production use.
We say no to work we have not shipped
We are deep in the Microsoft stack and current on the JavaScript one. If your problem needs something outside that, we would rather tell you on the first call than learn it at your expense.
Things we have built
Eleven products for one global marine logistics and ship-supply group
A nursery-software proof of concept turned into a multi-tenant SaaS
Cross-platform companion apps for environmental data loggers
Have something that needs building?
A thirty-minute call with an engineer, not an account manager. We will tell you if it is not work for us.