How we work

A delivery process your team can review.

A project should give you enough evidence to decide what to build, whether a release is ready and what your team will inherit. We agree those review points with you and keep decisions with the work they affect.

See our delivery process

What you review at each stage

We set checkpoints around decisions your team needs to make. The evidence can be a workflow map, an interface test or a release checklist, depending on the risk and the people responsible for sign-off.

Agree what the first release must do

Follow a representative case from intake to completion with the people who use and support the system. Include an exception and agree what must work on day one, which dependencies can block it and who can settle open questions.

  • Current workflow and priority user needs
  • First-release scope and acceptance criteria
  • Named decision owners and release dependencies

Test the architecture assumptions

Review data ownership, interfaces and hosting with your technical leads. Test a difficult assumption early, such as whether a vendor endpoint exposes the necessary records or can handle the expected volume.

  • System boundaries and data ownership
  • Architecture decisions with alternatives and trade-offs
  • Security, access and migration requirements

Build and release in testable increments

Give reviewers a complete task to try in each increment, including a representative failure or permission boundary. Use the results to decide what is ready for rollout and what would trigger recovery.

  • Versioned code linked to the work it implements
  • Test results for critical journeys, permissions and failure paths
  • Release notes and a rollback or recovery plan

Hand over the system in use

Walk through a deploy and a likely support incident with the people responsible for maintenance. Record where to check system behaviour, how to recover and which changes or risks carry into the next backlog.

  • Operating instructions and maintenance context
  • Documented ownership and escalation paths
  • An agreed backlog and support arrangement

Trace a requirement through to release

Before implementation we agree where requirements, implementation decisions and release evidence live. Your team can see the state of an item and follow its history without reconstructing it from meeting notes.

Azure DevOps for delivery records

When we use Azure DevOps, we link work items and acceptance criteria to the relevant pull requests and, where available, build or release records. Reviewers can follow what was agreed, what changed and what passed review.

Microsoft Teams for collaboration

When we use Microsoft Teams, we agree where to raise questions and how to involve the people who can resolve them. Decisions that affect the build go back into the relevant work item or decision record for future contributors.

Two ways to work with us

Project delivery suits a defined outcome with its own acceptance and handover. Embedded engineering suits an existing team and a shared backlog. In both cases we agree review, release and support responsibilities before work begins.

Project delivery

A good fit for a new portal, integration or phased replacement. Agree the intended outcome, interfaces, acceptance evidence and what your team will receive at handover.

Scope a project

Embedded engineering

A good fit when the team already owns the product and needs more delivery capacity. Agree where we can make changes, who reviews them and how we contribute within your planning and release rhythm.

Explore ongoing engineering