The challenge
A UK care home group runs several homes under more than one company. Every resident needs coded assessments, scheduled clinical charts and reviews that happen on time.
Each chart carries its own rules for targets, scores and escalation. Nurses must catch trouble early: fluid intake running low, days without a bowel movement, or malnutrition risk creeping up.
The group wanted care home care planning software shaped around its own forms and clinical rules. A generic package would have made its nurses work the way the software does.
It also wanted care recorded where it happens. Carers needed to log entries on the floor, while nurses needed a fuller view at a desk to compare, review and sign off.
The care home care planning software we built
We have worked with the group since 2023, building one module at a time from specifications written by its own team. The platform runs on React, Node.js, Express.js and MongoDB, with a mobile app that carers use on shift.
A form builder with AI-drafted care plans. Admins build coded assessment forms with review schedules, conditions and witness rules, without waiting for new development.
We connected the OpenAI API at three points. It suggests clearer question wording while a form is built, helps staff phrase their answers, and drafts the care plan summary. A nurse reviews, edits and acknowledges every draft before it stands.
Clinical charts with the rules built in.
Fluid targets are worked out from body weight, then refined from the resident's first two weeks of entries.
Bowel charts use the Bristol scale, and nurses are alerted after three days without a bowel movement.
MUST malnutrition scores and NEWS2 observation scores are calculated as readings are entered, with the next review proposed.
For behavior and enrichment charts, AI summarizes a chosen date range and reminds the nurse to check the source entries.
A body map. Every wound, pain point and skin mark sits on one color-coded map with its next review date. Each mark links to its own chart, and a painful wound opens the pain chart next.
Hospital passport and handover. One click builds a PDF hospital passport from the resident's profile, allergies, wounds, contacts and the uploaded MAR chart. The newest module, a shift handover for each wing, is built to show only the items that need attention.
How it works day to day
Follow a nurse through an early shift. When each resident was admitted, the platform switched on the charts and forms that resident needs. By the time the shift begins, carers find today's due charts on mobile and the nurse sees the same list on a desktop.
Carers record fluids, meals and checks as they go, and the scores calculate themselves.
When a rule is breached, such as three days without a bowel movement, a review task lands with the nurse. The nurse then compares recent entries side by side and can print a timeline. If a resident needs to go to hospital, the passport is a click away.
The handover module, the latest to be completed, is set up so the receiving nurse signs at changeover with a PIN. The handover then locks for audit. Other modules on the same platform cover staff compliance, resident admissions and allergies, meals and resident money.
The result
Carers record care on mobile and nurses review it on desktop. Admission switches on the charts and forms each resident needs, and they appear as due actions for the right role.
Rule breaches create review tasks instead of depending on someone spotting a pattern, and every AI draft is reviewed by a nurse before it counts. The group's clinical rules now live in the software rather than in individual memory.
What the research says
Most providers have already gone digital. By March 2026, 83.7% of CQC-registered adult social care locations in England were using a digital social care record, against 76.8% twelve months before. Those are Department of Health and Social Care estimates (GOV.UK).
Digital records give time back. The government estimates digital social care records give back at least 20 minutes per care worker per shift. They also let care plans be signed off in 3 days instead of 7 and reviewed in half an hour instead of 4 hours (GOV.UK, December 2025). Charting on a phone during the shift, with scores worked out as readings go in, is built for that kind of saving.
AI drafts can hold their own. In a 2024 Nature Medicine study, physicians rated summaries from the best adapted language models as equal to medical experts' in 45% of cases and better in 36% (Van Veen et al.). The same study logged errors on both sides, which is why a nurse reviews and acknowledges every AI-drafted care plan here.
Where AI fits next
Beyond the AI drafting already in place, these are ideas rather than features. AI could draft the shift handover from the day's chart entries for the outgoing nurse to check and sign. It could flag slow trends that span several charts, such as weight and food intake together, and compare wound photos over time to help nurses judge healing.
Planning something similar?
Before you build care planning software, it helps to answer four questions:
Which forms and charts are yours, and who owns their clinical rules?
Where should AI help, and who signs off what it drafts?
Which entries happen on the floor, and which at a desk?
How will you move from your current records without a gap in a resident's history?
Learn how we run custom software projects and AI automation, what we do about data privacy when AI is involved, and where else we have worked in health and care.
