Plain answers about Care-Meter for care home owners, managers and admins.

Care-Meter is the inspection-readiness platform for UK adult social care. It’s built for small and mid-sized care homes that are paper-reliant or running partial digital records — the homes most exposed to inspection risk.

It turns handwritten notes, voice memos, and photos into structured, CQC-defensible evidence without forcing your staff to abandon how they already work.

Pricing is flat per home, not per user. Your bill doesn’t change when staff turn over.

• Free — £0/forever. Up to 10 residents on Layer 1. No card, no time limit. Where most small homes start.
• READY — £149/month. Layer 1 with no resident cap, plus voice notes and unlimited carers.
• GOVERN — £249/month.Layers 1 + 2: adds the governance loop — manager review queue, Exception Inbox, Duty of Candour tracker, audit vault, 34-statement coverage view.
• INSPECT — £349/month. All three layers: adds PIR drafting, mock inspection across all 34 SAF Quality Statements, Reg 18 notification drafter, evidence export bundles, priority support and dedicated onboarding.
• GROUP — from £199 per home / month.2–10 homes, with a cross-home view and group dashboards.
• ENTERPRISE — negotiated. 10+ homes, with API access and SSO.

Annual billing saves 2 months on every paid tier.

Two different things. Free is forever— up to 10 residents on Layer 1, no time limit. It’s where most small homes start and many of them stay until they outgrow it.

The 14-day trial sits on top of paid plans (READY, GOVERN, INSPECT) so you can run a paid tier at full feature scope before committing. Cancel any time. GROUP and ENTERPRISE follow the same pattern: book a demo, run the trial, then commit.

In adult social care, staff turnover is high and admin capacity is low. Per-seat pricing punishes growth and creates friction every time someone joins or leaves. Per-home pricing is simple, predictable, and matches how owners and finance directors actually budget.

As little as we can avoid. Carers can keep writing notes by hand and photograph them at the end of a shift, dictate after handover, or type directly. We deliberately don’t ask staff to learn a new workflow on day one — the system meets them where they already work.

Yes. Every note — typed, photographed, or dictated — lands in a manager review queue with a status of PENDING_REVIEW before it becomes part of the permanent record.

You see what was captured, the AI’s reading of it, a fidelity score, and any clinical or safeguarding risk flags before signing off. Every approval, edit and override is written to the immutable audit vault.

On the INSPECT plan you get:

• A live coverage map showing where evidence is strong or thin across all 34 Quality Statements and the five Key Questions.
• A PIR drafting assistant that pre-populates the Provider Information Return from your live records.
• A mock inspection mode.
• A Regulation 18 notification drafter.
• Key Question-structured evidence export bundles for handing over to inspectors or commissioners.

The point is no three-day scramble before the inspector arrives.

• Regulation 17 — Good Governance. Contemporaneous records, manager review queue, immutable audit vault, evidence completeness checks.
• Regulation 18 — Notification of Incidents. Auto-classification of notifiability on every incident; a Reg 18 notification drafter on the INSPECT tier.
• Regulation 20 — Duty of Candour. Sequential Reg 20 compliance chain on every notifiable incident, with overdue alerts in the Exception Inbox.
• Single Assessment Framework. Live evidence coverage map across all 34 Quality Statements and the five Key Questions.
• Provider Information Return (PIR). AI drafting assistant on the INSPECT tier.

Yes — and the inspection-readiness layer on top.

Care-Meter holds the structured resident record that a DSCR is built from: residents, care plans, needs assessments, mental capacity assessments, advance care plans, risk assessments, daily care notes, incidents, and the audit trail behind every entry. Records are exported in a structure compatible with DSCR expectations.

Where Care-Meter differs from DSCR-only systems is what sits on top: a governance layer (manager review, Exception Inbox, Regulation 20 chain) and an inspection layer (PIR drafting, mock inspection, Key Question evidence export). The same record powers all three.

Yes — on the INSPECTplan. The PIR drafting assistant reads from your live evidence base and pre-populates a draft of your annual Provider Information Return, so you’re editing rather than starting from scratch.

We follow the current PIR structure as published by CQC; we don’t hard-code outdated formats.

When an incident is logged and classified as notifiable, Care-Meter opens a Regulation 20 compliance chain— verbal apology, written follow-up, investigation update — and surfaces overdue steps in the Exception Inbox.

The audit vault timestamps every step so you can show the full Duty of Candour sequence on demand.

All Care-Meter data is hosted on AWS in the UK / EU region (eu-west-2, London). Tenant data is isolated by your organisation ID and home IDat the API authorisation layer, so one home’s records cannot be queried by another.

Inside your organisation, access is gated by role: ADMIN, MANAGER, NURSE, SENIOR_CARER, CARER and FAMILY each see only what their role permits.

• Storage is encrypted and versioned in AWS S3.
• Every audit log entry carries a SHA-256 cryptographic hash, so any tampering is detectable.
• CI/CD deploys use OIDC role assumption— no long-lived AWS credentials in our pipelines.
• Authentication runs on AWS Cognito, with role-based access control enforced at the API layer.

Five — and we publish the list. Care-Meter is run with deliberately fewer sub-processors than is common in this category.

Resident data does not pass through marketing analytics tools, and we do not licence anonymised resident cohorts to research bodies or third parties. The full list lives at /sub-processorsand is updated whenever it changes — we mirror it from Schedule 3 of our DPA.

No. Care-Meter calls foundation models on AWS Bedrock in eu-west-2. Bedrock is configured so prompts and outputs are not retained for model training, and the underlying models are not fine-tuned on your data.

The specific models in production today, and what each is used for:

• Claude Sonnet 4.6 (eu.anthropic.claude-sonnet-4-6) — PIR drafting and the Copilot.
• Claude 3.7 Sonnet (anthropic.claude-3-7-sonnet-20250219-v1:0) — handwriting OCR and vision reasoning.
• Claude 3 Haiku (anthropic.claude-3-haiku-20240307-v1:0) — OCR fallback when Sonnet is throttled.
• Amazon Nova Pro (amazon.nova-pro-v1:0) — PIR drafting fallback.
• Amazon Nova Lite— Regulation 18 notification drafter.

Every AI-generated output is held in a manager review queue before becoming a permanent record, and every model call records the model identifier and prompt version for auditable provenance.

When a carer submits a note — typed, voice, or photographed handwriting — Care-Meter:

• Extracts the text (AWS Textract for handwriting, Amazon Transcribe for voice).
• Runs the extracted text through Claude on AWS Bedrock to produce a structured summary.
• Classifies it against the 34 CQC Quality Statements.
• Assigns a fidelity score.
• Flags clinical or safeguarding risk indicators.
• Routes the note to the manager review queue with status PENDING_REVIEW.

Every model call is logged with the model identifier and prompt version. The original capture (photo, voice file, or typed text) is preserved so you can re-verify against source.

Nothing reaches the permanent record without a human signing off.

Fidelity is our measure of how reliably the digital record reflects the original capture— particularly important for handwritten photos where OCR confidence varies.

A low fidelity score routes a note to closer manual review; a high fidelity score still goes through a manager but with a clearer signal that it can be approved quickly. It’s transparency about confidence, not a substitute for human oversight.

Three ways to correct it:

• Carersediting their own draft can correct or dismiss the AI’s suggested classification, summary, or extracted detail before submitting.
• Managers in the review queue can edit, approve, or return the note to the carerwith a written reason — for example “Please add the time this occurred and the outcome of the intervention.”
• If a verified note later needs a fix, an additional override is recorded as a discrete AI_VERIFICATION_OVERRIDE event.

The original capture (photo, voice file, or typed text) is preserved in every case so you can re-verify against source.

Every action — RETURNED_TO_CARER, AUTO_APPROVED, AI_VERIFICATION_OVERRIDE— is hash-sealed in the immutable audit vault with its own SHA-256 hash.

Admins manage staff inside the People tab in Settings. From there you can:

• Invite staff by email— invitations carry the assigned role and expire if not accepted.
• Assign roles — ADMIN, MANAGER, NURSE, SENIOR_CARER, CARER, FAMILY. Each role’s access is enforced at the AppSync API layer, not just in the UI.
• Edit a person’s role, deactivate them, or remove them when they leave.

All of this is self-service. Carers, nurses and senior carers can be added by an admin without contacting Care-Meter support.

Yes — on GROUP (2–10 homes) and ENTERPRISE (10+) plans.

Admins use the Homes tab in Settings to:

• Add a new care home to the organisation, including its CQC location ID and registered manager.
• Edit home details as registrations change.
• View the home directory across the organisation, with cross-home risk and evidence rolling up to the group dashboard.

Tenant isolation still applies inside the organisation: a home’s records are visible only to staff associated with that home, plus the organisation’s admins.

Note approvals. Admins set the home’s approval policy in the Approvals tab. Default policy is manager review on every note; admins can configure auto-approval thresholds (for example, notes above a fidelity threshold from senior carers) within the policy bounds Care-Meter publishes.

Feature scope. Features are gated by your plan tier (READY, GOVERN, INSPECT, GROUP, ENTERPRISE) via a single hasFeature(orgPlan, featureKey) rule the app enforces everywhere. Admins upgrade or downgrade their plan from the Billing tab; the change takes effect at the next billing cycle.

Sensitive admin actions — plan changes, role escalation, home creation — are written to the audit vault.

No.You start on Care-Meter with new notes from day one and digitise historical records on whatever schedule suits — typically by photographing them as they come up in resident reviews. Inspection-readiness compounds from the day you start, not the day you finish migrating.

For homes that want their backlog converted faster, we’re rolling out a digital-only Bulk Backfill Service:

• You scan or export your records yourself.
• Upload to a secure home-scoped intake folder. Supported formats: PDF, JPEG, PNG, TIFF, HEIC, CSV, XLSX, DOCX.
• Each file goes through the same OCR + Bedrock classification pipeline as live notes and lands in your manager review queue, with the original file preserved.
• Pricing is per page processed, on top of your plan.

Why digital-only?We never take physical custody of your residents’ records. That eliminates chain-of-custody risk, courier loss, and any secure-destruction obligation you’d otherwise be transferring to us. You keep your own paper; we work from your scans.

Care-Meter is built by Willmachi Limited (t/a Care-Meter), Co. No. 14540910, ICO registration ZC107807, based in Sittingbourne, Kent.

Support for all paying tiers is at support@care-meter.co.uk. GROUP and ENTERPRISE plans include onboarding support and priority response.