CalmPath is a closed-loop clinical system that turns routine dementia care into structured, learning evidence — matching non-pharmacological interventions to each person, correlating outcomes with PRN psychotropic use, and generating CQC-ready proof of person-centred care.

Antipsychotic and sedative overuse in dementia is an explicit patient-safety priority for the NHS Long-Term Plan and CQC — not a matter of clinical debate. Care providers currently have no tools to measure whether structured non-pharmacological care is actually reducing medication burden.
Agitation, restlessness or distress emerges — often at predictable times of day.
Frontline carers lack decision support, protocols and confidence in non-pharmacological options.
Antipsychotics or sedatives are administered off-label, at doses or durations not supported by clinical guidance.
1.5–2× stroke risk, 1.5–1.8× mortality risk, falls and sedation — with no data linking care quality to medication use.
Up to 70–80% of these prescriptions are at doses or durations not supported by clinical guidance — meaning 125,000–145,000 people are likely on inappropriate antipsychotic medication today.
A 2026 independent novelty search confirmed no existing solution combines the closed-loop personalised intervention engine, the Medication-Outcome Correlation Engine and integrated escalation-and-governance in a single platform.
Personalised protocol library, cognitive stimulation, digital life-story book, emotional wellbeing tracker and reminiscence aids — a Closed-Loop Intervention Engine that learns which non-pharmacological approach works for each person.
Structured daily morning check-in, decision support for UTIs, dehydration, delirium and falls, automated SBAR escalation and Constraint-Aware Predictive Risk Flagging — so no sign of deterioration is missed.
Video and phone companion sessions, a healing and relaxation hub, personalised wellbeing calendar and reminiscence content — accessible, person-centred engagement that combats isolation.
Just-in-time CQC, MCA and safeguarding guidance, carer wellbeing assessment, an AI care assistant for documentation, and a recognition system — reducing burnout and turnover.
Consent-controlled Family Portal, auto-generated CQC Evidence Dashboard and full audit trail — plus the Medication-Outcome Correlation Engine, which holds non-pharmacological outcomes and PRN psychotropic administration in one data model to generate the evidence clinicians need for safer, Clinician-led Medication Review.
Carer observes and logs a trigger — e.g. agitation at 3 pm.
System matches the trigger against the person's personalised behavioural profile.
Carer receives ranked, evidence-based non-pharmacological interventions and implements the best fit.
Structured outcome captured — distress before/after, engagement time, effectiveness.
Analytics update the model. Recommendation accuracy improves for that individual with every use.
The engine never alters prescribing. It generates the structured, auditable evidence that supports Clinician-led Medication Review and safer deprescribing — with clinical authority remaining with the doctor.
1950s jazz surfaced as Mr Brown's top-ranked intervention with 85% historical success; distress falls 4/5 → 2/5, and confidence in that recommendation strengthens.
The daily check-in flags new confusion and reduced appetite for Mrs Patel; decision support prompts a urinary check and auto-generates an SBAR handover and GP summary within minutes.
A GP reviews CalmPath's Medication-Outcome Correlation Engine dashboard: PRN antipsychotic use down 40% over 3 months alongside stable mood and engagement — evidence to support a safe dose reduction.
The registered manager pulls a real-time CQC Evidence Dashboard covering Safe, Effective, Caring, Responsive and Well-Led — ready in minutes, not weeks.
A daughter 200 miles away sees her mother's weekly mood trend and recent photos through the Family Portal — trust without another anxious phone call.
A declining pattern in a carer's daily check-in privately flags to the registered manager, who initiates a supportive conversation before burnout becomes turnover.
The MVP proves the innovation and delivers measurable value without over-building before customer validation. Success is gated on 3–5 onboarded pilot sites, ≥75% user satisfaction, consistent daily outcome capture and zero security or compliance incidents.
| Feature | Target month |
|---|---|
| Personalised dementia profile | Month 2 |
| Non-pharmacological protocol library (15–20 protocols) | Month 3 |
| Closed-loop outcome capture | Month 4 |
| Daily morning check-in | Month 4 |
| Clinical decision support (5 high-priority conditions) | Month 5 |
| Smart escalation with SBAR generation | Month 6 |
| Consent & Mental Capacity Act compliance framework | Month 6 |
| Basic analytics & CQC Safe-domain reporting | Month 7 |
| Capability | Generic care software | Remote monitoring | CalmPath |
|---|---|---|---|
| Personalised, ranked intervention recommendations | — | — | ● |
| Structured outcome capture per intervention | — | — | ● |
| Medication-Outcome Correlation Engine in one data model | — | — | ● |
| Automated SBAR generation & GP summary | — | — | ● |
| Standardised daily observation schema | — | ● | ● |
| Constraint-Aware Predictive Risk Flagging | — | — | ● |
| Continuous CQC evidence as a byproduct of care | — | — | ● |
A typical 60-bed care home spends £15,000–£25,000 annually on PRN psychotropic medication. A conservative 15–25% reduction in inappropriate PRN medication use — a core Business Objective — alone exceeds CalmPath's annual subscription within the first year.
Effective cost per person under care: approximately £25–£124 per year — negligible against £30,000–£50,000+ annual per-resident care cost.
MSc Biotechnology (Coventry, 2023). Level 5 Diploma in Leadership for Health and Social Care. 3+ years of senior care leadership managing rotas, CQC compliance, safeguarding and Mental Capacity Act 2005 practice. CalmPath is designed from CQC's five key lines of enquiry — not retrofitted to them.
B.Tech Computer Science & Engineering (D.C.R.U.S.T, 2018). End-to-end web development, front-end architecture and AI-assisted development. Prior B2B inside-sales development — an unusual combination that pairs technical delivery with a working understanding of customer acquisition.
A dedicated AI/ML engineer joins in Month 20 for later-phase predictive analytics and cross-cohort medication-correlation modelling. Phase 1–2 is deliberately scoped to rule-based, explainable logic appropriate for a patient-safety-critical context.
We're onboarding 3–5 pilot sites into the founding cohort with waived onboarding fees in exchange for structured feedback. Turn everyday care into the evidence CQC, families and clinicians need.