The operating layer for health systems moving care out of hospitals.
One open person record carrying clinical, operational and governance data, self-hosted, with the shift measured continuously rather than declared annually.
Three shifts are now national policy. None of them are measurable.
Hospital → Community
Care moves closer to home, but the record does not follow it.
Analogue → Digital
Records are digital; the data behind the decisions is still days or weeks old.
Treatment → Prevention
Prevention requires population-level targeting that operational systems cannot do.
Systems are mandated to shift care left. They cannot yet show whether it is happening, at what pace, or who is being left behind. Meanwhile the commissioning layer itself is moving — England’s integrated care boards merged from 42 to 26 on 1 April 2026.
Transformation is reported in board papers, not in data.
The record is fragmented
Acute, community, mental health, social care and primary care each hold a different version of the same person.
Progress is self-declared
Maturity is assessed by survey and narrative, months after the fact.
Equity is invisible
Aggregate improvement hides widening gaps in older and more deprived populations.
The scale of it: a two-million-population system runs 40 to 80 separate systems and employs 80 to 150 people whose job is moving data between them. That headcount is the integration layer, and it is the most expensive one available.
One platform. One person record. Everything else plugs into it.
LeftShift Intelligence™ — is the system actually shifting care left?Seeded demonstration
A single index for whether a system is shifting care left, with an equity modifier that penalises uneven progress — so a system cannot score well by improving only its easiest cohorts. The index is computed by the platform; the values shown are seeded demonstration data, not a deployment.
The index resolves to operational reality.Seeded demonstration
These are the measures the platform computes, at seeded values. There are no audited client outcomes yet, and we will not present these as though there were. First reference deployments are what turns them into audited ones.
Every number resolves to a named list of people.Seeded demonstration
Cohorts are derived from live population data, prioritised by urgency, and handed to the team that can act. Every tile drills through to exactly the set it counts, and the count reconciles — that is a platform standard, not a feature.
The board pack writes itself. A human signs it off.
From live operations
Pathway movement, areas requiring action and a drafted executive narrative, composed from the same record the wards are using.
Incident to objective
Event, theme, systemic issue, risk, board objective — one chain, evidence carried at every promotion, counts that reconcile.
Review before publication
Every AI output is held for approval. Sign-off is a recorded signature ceremony on a tamper-evident audit chain.
Non-SaMD by design: the platform surfaces recorded data and computed indicators for a human to act on. It does not diagnose, and it does not publish without a signature.
Configured per jurisdiction, not rebuilt per customer.
16 configured today
England, Scotland, Wales, Northern Ireland, Ireland, Crown Dependencies, Gibraltar, Jamaica, Antigua, Montserrat, Kenya, Oman, Saudi Arabia, United States. Policy, pathways and coding travel as configuration.
openEHR · SNOMED · FHIR
Vendor-neutral record, OMOP for secondary use, national canonical-model conformance. There is no lock-in argument to win because leaving is a designed property.
Non-SaMD by design
Intelligence and orchestration, not diagnosis. The boundary is asserted per module and reviewed at every regulated release.
One codebase, 613 modules, 570 live. Variance is modelled in a signed, versioned jurisdiction profile — never forked into local code, which is what makes the sixteenth jurisdiction cost the same as the second.
Land at a pound. Expand per employee. Scale per head.
A 12,000-staff trust is about £480k a year, against the £400–700k it already spends across eight separate governance products. UK governance and assurance software spend is £150–250m a year, £85–145m of it inside English trusts. Software gross margin above 80% — self-hosted, no per-seat infrastructure cost.Verified pricing card
Built by people who have worked inside health systems, not sold to them from outside.
FBCS · FIoD · CHCIO
Vice-Chair, HIMSS UK & Ireland
HSJ 100 most influential, 2025
Engineering, Azure and test leads on sweat equity. Chief executive hire is the first use of proceeds. Clinical advisory board on request.