An AI chatbot development for a UK care home handles the family enquiries and admissions information requests that currently require a care home manager or senior carer to step away from resident care to respond. For a UK residential, nursing, or dementia care home receiving 50-150 family and prospective resident enquiries per month, an AI chatbot handles 55-65% of routine enquiries automatically - room availability and fees, care specialism information, visiting hours, activities programmes, and CQC inspection report access - while ensuring all complex or distressing enquiries reach a human immediately. Setup costs £1,800-£4,500. Softomate Solutions builds CQC-compliant AI chatbots for UK care homes.
Last updated: 18 May 2026
Published 18 May 2026The UK adult social care sector entered 2026 in a state of sustained workforce crisis. Skills for Care reported 152,000 vacancy posts across adult social care in England in 2025 - a vacancy rate of approximately 8.3%. For individual care homes, this translates directly into a management team stretched beyond its operational capacity. Registered managers, deputy managers, and senior carers are simultaneously responsible for regulatory compliance, care quality oversight, staff supervision, and resident welfare. Every telephone call from a prospective resident's family - however routine the enquiry - represents a 10-15 minute interruption to that already-stretched capacity.
The volume of those interruptions is rising. Families now research care homes online before making direct contact. They arrive at the first telephone call having already compared CQC inspection ratings, read reviews on carehome.co.uk, and formed an initial impression of the home's specialism and fees. That first call is no longer purely exploratory - it is a confirmation conversation, often covering specific questions about room availability, fee structures, and the home's approach to dementia care or end-of-life care. The information required is largely factual, largely consistent, and largely already published on the home's website or in its statement of purpose.
The UK AI chatbot market reached £420 million in 2024 and is projected to grow to £1.1 billion by 2028 (CAGR 27%). UK businesses deploying AI chatbots report average first-response time reduced from 4 hours to under 10 seconds. Customer satisfaction scores (CSAT) for AI chatbot interactions average 3.8/5 in the UK, compared to 4.1/5 for human agent interactions - a gap that narrows to under 0.1 when the chatbot handles only in-scope queries. 78% of UK adults have interacted with a chatbot in the past 12 months; 54% prefer chatbot interaction for routine enquiries outside business hours. UK chatbot abandonment rate averages 35% when response time exceeds 10 seconds. AI chatbots reduce UK customer support costs by an average of £8-14 per ticket deflected (versus £12-18 for human agent handling). UK businesses with AI chatbots report 23% higher lead capture rates from website traffic versus businesses using only contact forms. GPT-4o API costs for a UK business handling 1,000 chatbot conversations per month average £40-80/month in API fees.
Yet those factual, consistent, already-published answers still require a human to deliver them. A care home receiving 80 family enquiries per month - a realistic figure for a 40-50 bed residential home in a competitive local market - is absorbing 800 to 1,200 minutes of management and senior staff time every month on information provision that could be automated. At a senior carer hourly rate of £13-16, that is £173-£320 in direct staff cost each month, before accounting for the care-quality cost of the interruption itself.
The communications pressure does not stop at admissions. Families of existing residents contact care homes with questions about visiting arrangements, activities schedules, GP appointment outcomes, and care review meeting dates. In homes without a dedicated family liaison role - the majority of UK residential care homes - these calls route to whoever is available, which is typically the same management team already managing regulatory demands.
An AI chatbot deployed on a care home's website and WhatsApp channel answers routine enquiries in under two minutes, 24 hours a day, seven days a week. A family researching care options at 10pm - a common pattern among adult children researching options for an ageing parent while their own children are in bed - receives an accurate, compassionate, factual response immediately rather than waiting for a callback during business hours. The home's management team starts Monday morning with one fewer urgent callback to make. That is the practical value proposition: not replacing human care, but protecting the human time that delivers care.
The AI chatbot's scope must be precisely defined before deployment. Scope creep into clinical or safeguarding territory is not a commercial question - it is a regulatory and ethical boundary. The following table sets out what an AI chatbot handles, what it routes to a human, and what it must escalate immediately without any AI involvement.
| Enquiry type | AI chatbot action | Notes |
|---|---|---|
| Room availability and current fees | Answers directly from published fee schedule | Fees are factual, not clinically assessed; weekly fee ranges are publishable without breaching privacy |
| Care type information | Explains residential, nursing, dementia, and respite care distinctions | Descriptive only - no clinical advice or assessment |
| CQC inspection report access | Provides direct link to the home's CQC profile | Mandatory to provide on request under CQC registration requirements |
| Visiting hours and policy | States current visiting policy including any infection control restrictions | Policy must be kept current in the chatbot knowledge base |
| Activities programme overview | Describes weekly programme categories and themed events | General overview only; specific daily timetable requires human confirmation |
| Viewing arrangement | Books a viewing via integrated calendar (Calendly or equivalent) | Confirms booking by email; human receives notification immediately |
| Trial stay enquiry | Acknowledges enquiry, captures contact details, routes to registered manager within 2 hours | Trial stays involve care needs assessment - human judgment required |
| NHS Continuing Healthcare enquiry | Provides factual summary of CHC process and routes to manager | CHC eligibility and funding are clinically determined - no AI assessment |
| Safeguarding concern | IMMEDIATE escalation - no AI response | Any mention of abuse, neglect, unexplained injury, or a vulnerable person at risk triggers immediate human contact and documented escalation log |
The safeguarding escalation row is non-negotiable. CQC inspectors specifically assess how care providers handle concerns raised by families and members of the public. An AI chatbot that attempts to handle a safeguarding concern - even with the most carefully worded response - represents a regulatory and reputational risk that no care home should accept. The correct architecture routes any message containing safeguarding-related language (including terms like 'bruise', 'not eating', 'frightened', 'hurt', 'unhappy', 'locked', 'neglect') to a human immediately, logs the conversation in full, and triggers a notification to the registered manager regardless of the time of day.
Beyond the safeguarding boundary, the chatbot's knowledge base is built from the home's statement of purpose, its published fee schedule, its CQC registration information, and its activities programme. Updates to visiting policy, fee changes, or activities schedule changes require the knowledge base to be updated - a process that takes minutes but must be part of the home's operational discipline. Softomate builds a simple content management interface into every care home chatbot deployment so that the registered manager or administrator can make updates without technical assistance.
CQC Regulation 17 (Good Governance) requires registered care providers to maintain effective systems and processes for assessing, monitoring, and improving the quality and safety of the services they provide. Communication systems - including digital communication tools - fall within the scope of that regulation. A care home deploying an AI chatbot is not operating outside regulatory oversight; it is operating a communication system that CQC inspectors may examine during an inspection.
Understanding which Key Lines of Enquiry (KLOEs) are relevant to AI-assisted communication helps care homes deploy these tools in a way that strengthens rather than complicates their inspection evidence.
CQC's 'Caring' line of enquiry assesses whether staff treat people with compassion, kindness, dignity, and respect, and whether information and support are provided to people and those close to them. An AI chatbot visible on a care home's website makes information accessible to families at any hour - an accessibility improvement that inspectors can see directly. The chatbot's tone must be configured to reflect the home's values: warm, clear, not clinical, not transactional. Softomate's care home chatbot template is built on a tone framework specifically designed for adult social care settings - it does not use customer service language that would feel jarring to a family researching care for a parent with dementia.
The 'Responsive' line of enquiry includes how people and those close to them can access information about the service and how concerns or complaints are handled. An AI chatbot directly addresses the accessibility criterion: it makes information available 24 hours a day in plain language, with the option to escalate to a human at any point. Critically, every chatbot conversation is logged with a timestamp, the enquirer's identifier (phone number or email if provided), and the full conversation transcript. That log is the evidence base that demonstrates to a CQC inspector how enquiries were handled - something that a telephone-based enquiry system cannot easily provide.
The 'Well-Led' line of enquiry examines whether the provider has effective governance arrangements that drive continuous improvement. A care home that can show a CQC inspector a dashboard of enquiry volumes, response times, escalation events, and knowledge base update logs is demonstrating exactly the kind of systematic governance that supports a 'Good' or 'Outstanding' rating in this domain. Most care homes currently have no structured record of family enquiry handling. An AI chatbot, paradoxically, creates better governance evidence than the traditional phone-and-notebook approach it replaces.
CQC inspectors specifically test how providers handle safeguarding concerns raised outside of normal working hours, or through informal channels such as website contact forms. During inspections, inspectors have been known to send test enquiries through digital channels to assess the provider's responsiveness. A care home whose AI chatbot routes safeguarding language to a human within minutes, generates a timestamped log, and notifies the registered manager immediately will pass that test. A care home whose chatbot attempts to manage the concern with a pre-written response will not. Softomate's implementation requires a safeguarding escalation review and sign-off before any care home chatbot goes live - this is not a feature that can be deferred or disabled.
It is worth noting that CQC does not prohibit the use of AI-assisted communication tools in registered care settings. The regulatory requirement is that communication systems - regardless of technology - must be compassionate, accessible, and properly governed. An AI chatbot that meets those requirements is compliant with CQC's framework. One that does not is not rendered compliant by being digital; it is rendered non-compliant by failing the governance standard.
The admissions enquiry use case is the most immediately visible application of AI chatbot technology in care homes. But for a home that has deployed an AI chatbot well, the greater long-term value lies in automating the ongoing communication with families of existing residents. That communication is currently one of the most time-consuming and least systematised aspects of care home operations.
A family with a parent in a 40-bed residential care home expects, at a minimum, to receive updates about significant events in their relative's care: GP visits, hospital appointments, falls (which must be reported under duty of candour), changes to care plans, and care review meeting invitations. In practice, many families also want to know about activities their relative has participated in, whether they have been eating well, and when other family members have visited. Delivering all of that information by telephone is not feasible at scale. An AI-assisted family communication system handles the routine notifications automatically while preserving human contact for the genuinely significant conversations.
Resident health and care data is special category personal data under UK GDPR, as defined in Article 9 of the retained EU Regulation 2016/679 as it forms part of UK law under the Data Protection Act 2018. That classification has direct implications for family communication automation.
The legal basis for processing a resident's health data in the context of their care is typically Article 9(2)(h) - processing necessary for the provision of health or social care treatment. But communicating health data to a family member - even a close family member - requires either the resident's explicit consent, or a legal authority to act on their behalf. For residents who lack capacity to give consent, that authority is typically held by the donee of a Lasting Power of Attorney (LPA) for health and welfare, or in some cases by a court-appointed deputy.
A care home deploying AI-assisted family communication must therefore verify, for each resident, the identity and legal authority of the person designated to receive health-related communications. This is not a technical challenge - it is an operational and legal discipline that the AI system supports by maintaining a consent register linked to each resident record. Softomate builds the GDPR consent management interface into every care home deployment, with LPA status fields and designated contact verification as standard. The ICO's guidance on special category data in health and social care settings - available at ico.org.uk - is the definitive reference for care homes assessing their data protection obligations in this area.
Communications that do not involve health data - visiting booking confirmations, activity notifications without health content, general home news - can be sent to any family member the resident or their representative has designated, without requiring a special category legal basis. The AI system distinguishes between these categories automatically, routing health-related communications only to verified designated contacts.
Softomate Solutions implements AI chatbots for UK care homes through a structured six-stage process designed specifically for the regulatory and operational environment of adult social care. The implementation is not a generic chatbot deployment with a care home theme applied; it is a purpose-built system with safeguarding, CQC governance, and GDPR compliance built into the architecture from the outset.
Before any other configuration work begins, Softomate works with the home's registered manager and safeguarding lead to design and document the safeguarding escalation pathway. This covers the language triggers that activate immediate escalation, the notification recipients (registered manager plus nominated deputies), the documentation format for escalation logs, and the out-of-hours contact protocol. The registered manager signs off the escalation design in writing before development proceeds. This is non-negotiable. No care home chatbot goes live without a documented, manager-approved safeguarding escalation protocol.
The knowledge base is constructed from the home's statement of purpose, its current fee schedule, its CQC registration profile, its activities programme, and its visiting policy. Softomate's content team structures this information into the chatbot's retrieval architecture - not as a simple FAQ list, but as a semantically organised knowledge graph that allows the chatbot to answer compound questions accurately. A family asking 'Do you have a room available for my father who has vascular dementia and needs nursing care, and what would it cost?' receives a single coherent response drawn from multiple knowledge base elements, not three separate answers.
The admissions enquiry flow is configured to the home's specific process: initial enquiry, viewing booking, needs assessment arrangement (human-led), fee agreement, and move-in coordination. The AI handles the first two stages; all subsequent stages route to the registered manager or admissions coordinator with a full conversation summary. The viewing booking integrates with the home's existing calendar system - Calendly, Google Calendar, or Microsoft Bookings.
Where the home wants to deploy the AI chatbot beyond admissions, Softomate configures the family communication workflows described in Section 4: visiting booking, activity notifications, care review invitations, and concern flagging. Each workflow is tested against the home's operational schedule and staff capacity before activation.
The consent register is built to the home's specific resident population, with LPA status fields, designated contact verification, and communication category permissions (health data, general communications, activities). The system generates a GDPR-compliant consent record for each resident that the home can produce for a CQC inspector or an ICO audit.
Softomate configures the chatbot's reporting dashboard to produce the evidence outputs most relevant to CQC inspection: enquiry volume by month, response time distributions, escalation event log, knowledge base update history, and conversation quality samples. The registered manager receives a monthly summary report by email. The dashboard is accessible to the CQC inspector on request during an inspection.
Implementation cost for a UK care home AI chatbot is £1,800-£4,500 depending on the scope of the deployment (admissions only, or admissions plus family communication workflows), the complexity of the home's care specialism mix, and the level of GDPR consent management required. Implementation takes three to six weeks from signed agreement to go-live.
Softomate recommends that care homes due a CQC inspection within six months schedule go-live after that inspection. A newly deployed system adds a governance variable to the inspection period; it is better deployed after inspection and allowed to mature before the next inspection cycle. Homes not in an active inspection window can deploy at any point in the year.
Ongoing support is included for the first six months: knowledge base updates, safeguarding trigger review, and GDPR consent register maintenance. Annual support packages are available from month seven.
Yes, when implemented correctly. CQC Regulation 17 (Good Governance) requires that communication systems are effective and properly governed - it does not specify that those systems must be human-only. An AI chatbot with full conversation logging, safeguarding escalation protocols, and documented management oversight meets the governance standard CQC inspectors assess. The key requirement is that the system is demonstrably managed and that safeguarding concerns are never handled by the AI.
The safeguarding escalation triggers an immediate notification to the registered manager's personal mobile via WhatsApp and SMS, plus a backup notification to the nominated deputy manager. Care homes must designate at least two recipients for overnight escalations. Softomate builds the notification routing to the home's specific on-call rota. The escalation log is timestamped and preserved automatically - the registered manager sees the alert and the full conversation transcript simultaneously.
Resident health data is special category personal data under UK GDPR and the Data Protection Act 2018. Communications that include health information may only be sent to a person with legal authority to receive it - typically the resident (if they have capacity), or the donee of a Lasting Power of Attorney for health and welfare. Softomate builds a consent register into every care home deployment that records each resident's designated contact and their legal authority. The ICO provides detailed guidance on special category data in health and social care settings.
For a 20-bed residential care home deploying an admissions-focused chatbot without the full family communication workflow, implementation typically falls at the lower end of the £1,800-£4,500 range - commonly £1,800-£2,400. The knowledge base for a smaller home with a single care specialism is simpler to build, and the enquiry volume is lower, reducing the configuration complexity. Ongoing support costs are the same regardless of home size, as the safeguarding and governance infrastructure requirements are identical.
The AI chatbot provides factual information about the NHS Continuing Healthcare process - what it is, how eligibility is assessed, and how the care home supports the assessment process - and then routes the enquiry to the registered manager. CHC eligibility is determined by a clinical assessment that requires qualified professional judgment; the AI does not attempt to assess eligibility or advise on likely outcomes. This is both a regulatory requirement and a practical protection for the care home against liability for inaccurate CHC guidance.
The chatbot is configured to detect distress signals - urgent language, emotional distress indicators, expressions of crisis - and route immediately to a human, with a warm message acknowledging the person and confirming that a member of the team will be in contact within a specified time. The system does not attempt to provide emotional support or counselling. Where the distress is connected to a concern about a resident's welfare, the safeguarding escalation triggers simultaneously. Softomate's tone framework for care home chatbots is specifically designed to handle distressed enquirers with compassion and clarity rather than transactional efficiency.
Well-configured AI chatbots handle 65-80% of UK website enquiries without human intervention. The remaining 20-35% are escalated to human agents due to: complexity beyond the chatbot's training data (typically 15%), explicit requests to speak with a person (typically 10%), and technical failures (typically 5%). UK businesses in sectors with highly standardised enquiries (dental appointment booking, trade quote requests, property viewing scheduling) achieve automation rates above 80%. Complex B2B sales queries and regulated advice requests (legal, financial, medical) are designed to escalate directly to humans.
UK care homes face a communication gap that is directly costing management time and indirectly affecting the quality of resident care. An AI chatbot handles 55-65% of routine family and admissions enquiries automatically, with every safeguarding concern escalated to a human immediately and every conversation logged as CQC-ready governance evidence. Skills for Care's 2025 figure of 152,000 adult social care vacancies confirms the workforce pressure is structural, not cyclical - care homes need communication systems that protect the human capacity that delivers care. Softomate Solutions builds CQC-compliant AI chatbots for UK care homes at £1,800-£4,500, with safeguarding escalation reviewed and signed off before go-live.
Explore Softomate's AI chatbot development service or contact us to discuss your care home's specific requirements.
Written by Rakesh Patel, Softomate Solutions, Barking, East London.More from the blog
Let us help
Talk to our London-based team about how we can build the AI software, automation, or bespoke development tailored to your needs.
Softomate Solutions Limited
Deen Dayal Yadav
Online
