How we work

Embedded squads. Shared ownership. Governed outcomes.

Not consulting

We do not hand over reports and leave. Clinicians, safety officers, and operational leads sit inside the team as co-owners.

Parallel tracks

Discovery and engineering run together from day one. The gap between understanding and building never opens.

Shown, not described

Every two weeks the squad demonstrates working output. Not slides. Not status reports. The actual thing.

Packaged at every boundary

The discipline is toward a complete, usable deliverable at every sprint boundary — not accumulated work-in-progress.

We are not a consulting engagement

The conventional model creates a gap between the people who understand the problem and the people who build the solution. In safety-critical healthcare, that gap is where governance is lost.

Conventional model

External discovery team arrives, conducts research, writes findings

Report handed to a separate delivery team who never met the clinicians

Clinicians become indirect inputs — quoted in documents, absent from decisions

Governance degrades in translation — the workaround described in week two is forgotten by week twelve

Inference Clinical

One embedded squad — service designers and engineers side by side from day one

Clinicians inside the team — a CSO who sees governance decisions as they are made, not after

Discovery and engineering stay coupled — the assessment produces working specifications that engineering acts on

Governance decisions remain live — in the room, not in a requirements document

How the work moves

One repeating cycle. Every engagement, every stage, every sprint. The cadence is the same whether you are in the assessment, the lighthouse, or scaling across boundaries.

01

Squad convenes

Service designers, engineers, and your clinicians and operational leads work as one team. Sprint scope is set from the backlog.

02

Discovery + engineering in parallel

Service design maps the choreography while engineering builds against emerging specifications. Neither waits for the other.

03

Packaged deliverable

The goal is a complete, usable output the organisation owns. Some sprints are research-heavy or hit blockers — the discipline is always toward packaging what has been learned.

04

Show and tell

The squad demonstrates working output to the wider organisation. Clinicians confirm. Safety officers validate. The organisation decides.

05

Retrospective

What the team learned changes the next sprint. The specification evolves. The work stays aligned with clinical reality.

Cycle repeats every two weeks

Show and tells, not steering committees

Every two weeks, the squad stands in front of the work and shows what it does. Not a slide deck. Not a status report. Working output, demonstrated to the people who will govern it.

Demonstrated

Mapped choreography

The handover pathway as it actually operates — not how the policy document says it should. Clinicians see their working reality reflected in the model.

Demonstrated

Governed boundary in test

A boundary running with supervision rules, responsibility transfer logic, and evidence recording. Working, not proposed.

Demonstrated

Supervision rules against real scenarios

Safety rules firing against real clinical scenarios. Safety officers confirm the governance model matches the regulatory obligation, in the room, not after the fact.

If something is not working — if the choreography is wrong, if the scope needs to change, if the boundary is not the right one to start with — the show and tell is where that surfaces. Not six months later in a programme review.

Packaged deliverables at every sprint boundary

The discipline is toward a deliverable that is complete in itself at every sprint boundary. Not every sprint achieves this — some are research-heavy, some hit genuine blockers. But the operating principle is always: package what has been learned, not accumulate work-in-progress.

Assessment sprint

Handover map

A validated map of one segment of the pathway — handover points, responsibility gaps, and governance constraints documented and confirmed by clinicians.

Example: Discharge pathway from acute ward to community nursing — seven handover points mapped, three governance gaps identified.
Lighthouse sprint

Governed boundary in test

One boundary running with supervision rules, responsibility transfer logic active, evidence being recorded. Proven before the next boundary begins.

Example: Ward-to-community boundary governed with bilateral acceptance, clearing rate measured at 94%.
Platform sprint

New boundary specification

A new boundary composed and ready for deployment. Each boundary builds on the infrastructure proven by the lighthouse — complexity managed one crossing at a time.

Example: Community-to-GP boundary specification composed, ready for deployment on existing SafeMesh infrastructure.

The principle: if the engagement paused tomorrow, the last packaged deliverable should still be usable. That is not always perfectly achievable — some work genuinely spans multiple sprints. But the discipline of packaging at boundaries means the organisation is never left holding only work-in-progress.

Learning at sprint ends

Every sprint ends with a retrospective that changes something. Not a ritual. Not a form. A genuine examination of what the team learned — and a specific commitment to act on it.

In healthcare delivery, this matters more than in most domains. The assessment will reveal things nobody expected. The lighthouse will encounter handover conditions that were not in the specification. The clinical context will shift. The squad absorbs those changes at sprint boundaries.

If this sounds like how you want to work, we should talk.

We work with NHS organisations, independent providers, and insurers who want an embedded team, not an external report.

Talk to us