Manifesto

Preamble

Every day, in more than half a million homes across the United Kingdom, a person arrives to care for another person. What happens in that hour is among the most consequential work a society performs — and among the least measured. We measure whether the visit happened. We measure whether the form was filled. We measure, after the fact, whether anyone complained. We do not measure the thing itself: whether the care was good, and whether it was theirs — fitted to the one person receiving it.

This is not because quality is unknowable. Every experienced care manager carries the knowledge in their head: who is qualified for whom, who is exhausted, who has the bond with Mrs. Adeyemi that took eight months to build and would take one careless rota change to break. The knowledge exists. It has simply never been formalised, and so it cannot be protected, scaled, or proven. When the care manager leaves, it leaves with them. When the schedule is rebuilt — as it is, weekly, in four out of ten home care organisations — it is rebuilt blind.

We founded Aultrum on a single conviction: quality and personalised care is not a sentiment. It is a structure — and structures can be modelled, measured, and defended.

The claim

We claim that the quality of care delivered by one human being to another is determined by six measurable layers, ascending from capability to meaning. We did not invent these determinants; care research, psychology, and law discovered them separately over decades. Our contribution is to unify them into one computable model — the Aultrum Layers — and to build the system that brings them to life: Persona Intelligence.

We hold each layer to a standard: it must be anchored in an established legal, clinical, or psychological framework. We refuse to score human beings against criteria invented in a product meeting.

I. Competence. Care begins with capability — not credentials held, but skills matched to this client's needs. Our anchors are the frameworks the sector itself built: the Care Certificate, the Regulated Qualifications Framework, the Core Skills Training Framework, and NICE guidance for the conditions that shape each life. A fully qualified carer can still be the wrong match. Competence is a relationship between a skill set and a person.

II. Compliance. Safety is not a layer that can be traded against the others; it is a gate. Our anchors are the law itself: the CQC fundamental standards under the Health and Social Care Act 2008 regulations, the Disclosure and Barring Service, the Mental Capacity Act 2005, and the UK GDPR. In our model, a single failed check zeroes the entire match. No degree of skill or warmth compensates for an unsafe placement. Ever.

III. Shift Readiness. A competent, bonded carer delivers degraded care when exhausted — fatigue science has shown this in every safety-critical profession, and care is a safety-critical profession. Our anchors are fatigue-risk-management research and validated instruments including the Maslach Burnout Inventory and the Utrecht Work Engagement Scale. We hold that protecting the carer's capacity to care is protecting the client, and we refuse the false economy that treats workforce wellbeing as separate from care quality.

IV. Rapport. This is the heart of the model and of personalised care. Rapport is not magic and it is not luck; psychology decomposed it decades ago. Tickle-Degnen and Rosenthal showed that rapport is built of mutual attentiveness, positivity, and coordination — observable components, and what can be observed can be understood. Bordin showed that the working alliance — bond, shared goals, shared tasks — predicts outcomes across the helping professions. Kitwood taught the care sector that personhood is maintained or destroyed in the small moments of relationship, and Brooker's VIPS framework made his philosophy practicable. Mercer's CARE measure proved relational empathy can be reliably assessed. Standing on all of this, we make our most original claim: the bond between a specific carer and a specific client can be respectfully measured, longitudinally tracked, and deliberately protected. No system on earth currently does this. We will.

V. Satisfaction. Outcomes belong to two people, not one. We measure whether care worked for the client — anchored in ASCOT, the Adult Social Care Outcomes Toolkit that UK commissioners already trust — and whether it worked for the carer. A model that optimises client satisfaction by consuming its workforce is not a quality model; it is an extraction model wearing one.

VI. Purpose. At the top of the hierarchy sits meaning. Self-Determination Theory — autonomy, competence, relatedness — explains why people stay in hard work or leave it, and Skills for Care's own evidence confirms it in our sector: carers with stability, training, qualification, and fair conditions leave at a third of the rate of those without. Purpose is where quality and retention become the same number. A workforce that finds meaning stays; a workforce that stays builds rapport; rapport is personalised care. The hierarchy closes its own loop.

The principles of Persona Intelligence

The Layers produce Persona Intelligence: a living, evidence-backed understanding of every carer, every client, and every bond between them. Power over such knowledge demands principles, and we state ours as commitments, not aspirations.

Quality multiplies; it does not average. Our match score combines the layers multiplicatively, so that no weakness can hide behind a strength. A brilliant bond does not excuse dangerous fatigue. This is mathematics in service of a moral position.

Continuity has a price, and we will charge it. Every rota system on the market will break an eight-month bond to save eight minutes of travel time, because the bond appears nowhere in its objective function. In ours, the destruction of relational capital is a cost, calculated and visible. We call this the continuity value, and we believe pricing it is one of the most humane things a mathematical model has ever been asked to do.

The Persona belongs to the carer. The Aultrum Persona is the carer's verified record of their quality — and it is theirs: portable across employers and platforms, carried with consent, built to make a career in care accumulate value the way careers in other professions do. We measure people for them, never merely about them. A care workforce that owns its own evidence is a care workforce with power.

Nothing about a person leaves their home. Ambient understanding is processed on the device, in the moment. Recordings are not stored. Only layer-level scores travel, and the models improve across the whole network through federated learning — every agency makes the system wiser, and no central archive of vulnerable people's lives ever exists, anywhere, including with us. Privacy is not our policy; it is our architecture, and architecture cannot be quietly amended.

Every judgement shows its working. Any score, prediction, or assignment our system produces can be decomposed back to its evidence — the compliance state, the competence fit, the readiness index, the rapport history. We build for the inspector, the family member, and the carer who asks why. An intelligence that cannot explain itself has no place in the care of human beings.

Autonomy is a dial, never a default. Our intelligence can inform, recommend, act with approval, or act alone — and the human organisation chooses, task by task. The judgement of an experienced care manager is not a inefficiency to be automated away; it is the expertise our system exists to amplify. And whatever the dial's setting, accountability remains human.

The evidence rules us. Every layer is a hypothesis as well as a conviction, and we will test our model against real outcomes, publish our methodology, and rebuild any layer the evidence defeats. We anchored ourselves in research because research submits to correction. So will we.

What we refuse

We refuse to optimise coverage and call it care. We refuse to surveil clients and call it safety, or to score carers in secret and call it quality. We refuse to let the intimate data of vulnerable people become anyone's asset, including ours. We refuse the resignation — comfortable, universal, false — that says care quality is too human to measure, a resignation that has protected nothing except the status quo that fails carers and clients alike.

The horizon

Demographics are not negotiable: this country will need almost half a million more care workers within fifteen years, while the supply of people willing to do undervalued, unmeasured, unprotected work contracts. The sector will be transformed — by intelligence, by agents, in time by robotics. The only question is what that transformation optimises. If the machines that coordinate care inherit today's objective functions, they will deliver coverage without quality at unprecedented efficiency. We are building the alternative inheritance: a formal, humane, auditable definition of what good care is — so that everything that comes after, human or machine, optimises for the right thing.

Quality care is computable. Personalised care is protectable. The people who deliver it deserve to own the proof.