We build applications that follow the way your teams already work — the approvals, exceptions and handovers that off-the-shelf products force you to work around. Every engagement produces source code you own, a data model you can read and a deployment path your staff can run.
Our builds fall into four categories, and most engagements combine two of them.
Internal systems for scheduling, job costing, asset registers and case management. Browser-based, role-aware, and built for all-day use by people who are not software specialists.
Interface approach →Offline-first mobile applications for crews working beyond reliable coverage. Jobs and inspections are captured on device, queued locally and synchronised when signal returns, with conflict rules agreed up front.
Mobile stack →Request, review and approve chains with delegation, escalation timeouts and a complete audit record, so an auditor can trace who approved what, when and on whose authority.
Related work →Replacing an ageing database tool or an unsupported line-of-business product with a maintained equivalent — staged so the old system keeps running until the new one has been reconciled against it, record for record.
Read our view →These are the situations that prompt a first conversation. The work gets done, but the cost sits in people's time and nobody can say with confidence that the numbers are right.
One workbook holds pricing, another holds allocations, and a third is emailed for sign-off. There is no audit trail, no concurrency control, and only one person understands the formulas.
A job is raised in one system, re-keyed into a second and reconciled against a third at month end. Every re-entry risks a transposed figure, and the reconciliation consumes days.
An off-the-shelf platform covers eighty per cent of the requirement; the rest is absorbed by workarounds and shadow spreadsheets. The licence is only part of the real cost.
Dockets and timesheets are completed on paper, then entered into the office system on Monday. Billing is delayed, discrepancies cannot be investigated, and nothing is queryable until it is typed in.
We do not report progress as a percentage. Each stage ends with an artefact you can examine: a model, a running screen, a migration dry run or a signed-off runbook.
Before interface work we map the entities your operation uses — sites, assets, crews, orders, lots — and the states they move through. Scope disputes are usually vocabulary disputes, so the glossary is signed by both sides.
Screens are designed around the task, not the database table. We prototype the highest-volume journeys, test them with daily users, and revise before production code is written.
Delivery runs in fortnightly increments against a deployed review environment, with automated tests on the rules governing money, stock and compliance.
Where an existing system is replaced we map it field by field, run dry migrations into staging and reconcile record counts. Cutover is scheduled in AWST with a rollback point.
Launch is a milestone, not the end of the engagement. We run handover sessions with your staff, leave runbooks and a support channel, and continue on a fortnightly rhythm for clients wanting further development.
The decisions that determine whether a system is still comfortable to change in three years are made in the first month. These are the ones we make deliberately.
Typical of the work described on this page. The client is identified by sector only.
A national wholesale distributor ran each of its four warehouses on a separate copy of the same spreadsheet, consolidated nightly by a supervisor working after hours. Discrepancies surfaced when a customer queried a short delivery rather than when the error occurred. We modelled stock at lot and location level, built a browser-based receiving and picking application with handheld scanning, and encoded the reconciliation rules that run unattended each evening.
Engagements are quoted against a written functional scope; exploratory work begins with a short paid discovery.
Hosting is billed separately and sits in your own cloud account. Ongoing operation is arranged under the maintenance service rather than bundled into the build.
Indicative only, drawn from engagements delivered since 2021. Focused internal tools sit below this band; multi-site platforms with mobile capture and migration sit toward the upper end. We issue a costed range after discovery, not a single number revised later.
Yes. The repository, schema, infrastructure definitions and documentation are transferred to you. There is no proprietary runtime, no per-seat licence payable to us, and no barrier to another firm maintaining the code.
Usually yes, and most of what we build sits beside an existing platform rather than replacing it. We design around what the system exposes — REST API, database view, export file or SFTP drop. Where a vendor API is thin, we say so during discovery.
They usually do. Changes are written down, costed and approved before work begins, and you see the effect on budget and schedule before deciding.
The same engineers who built it. Every build includes thirty days of defect rectification at no additional charge, and most clients continue into a support arrangement with defined response targets.
A 45-minute scoping call, no obligation. We will tell you honestly whether a custom build is the right answer.
