Iris is an AI companion and care assistant for older people in the UK, delivered on a dedicated tablet that works the moment it is plugged in. The label on every device says one thing: “Just say Iris to start talking.” No login. No menus. No learning curve.

Iris serves two markets, sequenced. B2B first: care homes subscribing on behalf of their residents, where Iris is the first product purpose-built to generate CQC evidence as a byproduct of resident companionship. B2C second: families subscribing directly for an elderly relative still living at home, a larger but harder market that the B2B installed base will seed.

The product is built on a care-specific orchestration layer sitting on top of frontier language models — proprietary prompts, per-resident memory, clinical signal logic, and care documentation pipelines. The models are not ours. The system of record built around them is.

Voice is not a new interface for older people. It is the original one. The care tech graveyard is full of tablets handed to residents and never touched again, because they were designed for users who learn new interfaces easily — users this generation is not.

Iris combines voice with a stripped-back touch layer. Voice mirrors conversation, the social skill that persists longest as cognition declines. Touch offers a visible fallback with no hidden menus and nothing to remember. For residents with dementia specifically, this matters: Iris does not require memory of previous sessions, does not require navigation, and is designed so the resident never feels they have got something wrong.

This is the structural departure from existing care technology. Monitoring tools serve the operator at the expense of resident privacy. Communication apps serve families but add work for carers. Documentation software serves compliance but does nothing for the person in the bed. Iris is built from a different starting point: the resident interaction is the primary data source, and every other stakeholder — carer, manager, family — is served downstream of it.

When a care home or family subscribes, a dedicated Iris tablet arrives pre-configured and branded. Setup is done once, by the people who know the resident best. The resident receives a device that already knows their name, their family contacts, and the basics of their life.

The hardware is a service, not a purchase. Iris owns and maintains every device. Replacements ship within 48 hours. At end of subscription they are collected. There is no procurement, no IT burden, and no capital expenditure for the subscriber.

Privacy and security are architectural, not configurable. All audio processing happens on-device until the wake word is detected. No raw audio is ever transmitted or retained. Conversation data is encrypted in transit and at rest, held in segregated per-resident environments in AWS London, under UK data protection law.

Iris is built to learn and remember a specific person over time. Every conversation refines what the system knows about them: their preferences, their sensitivities, the topics that bring them to life, the ones that cause anxiety, the patterns that signal something might be wrong.

The clinical layer runs on the same conversational substrate. The system that holds a conversation also listens for signals — changes in response patterns, mentions of physical discomfort, deviation from an established baseline — and surfaces them as structured alerts for carers. Companion and clinical tool are not separate features. They are the same pipeline, doing two things at once.

Iris serves four people simultaneously, each with a different relationship to the same resident. This is the hardest thing to build and the main reason existing products fall short: each was architected around a single buyer, and the other stakeholders inherit the compromise.

The resident gets a companion who already knows them. The carer gets AI-drafted care notes that take 90 seconds to review and approve instead of 45 minutes to write from memory. The manager gets continuously-generated CQC evidence, real-time operational visibility, and a retention tool in a market where carers leave because of paperwork. The family gets daily visibility without relying on carer workload: a warm note about how the day started, photos woven into conversation, video calls their relative can answer by saying yes.

One ecosystem. Four genuine wins. At a price that costs the home less than one hour of agency care per resident per week.

The proving ground. Care homes subscribe per bed per month, billed to the operator. Iris integrates with the care team's dashboard, generates documentation and alerts from day one, and delivers measurable carer time recovery and CQC evidence within the first 30 days of deployment. Pilot structure is designed to convert evidence into contracts: 30 days free, up to 10 residents, no commitment. This is where Iris builds the clinical track record, the case studies, and the operational maturity that make B2C credible at scale.

The expansion. The 85+ population will nearly double by 2047, and the majority will remain at home for as long as possible. The gap between that aspiration and the reality — isolation, missed medication, unnoticed deterioration — is where the B2C opportunity lives.

The B2C market is structurally larger than B2B, with a consumer subscription model, lower CAC through family referral, and higher gross margins. But it requires the clinical credibility and product maturity that only B2B deployment can produce. Iris launches B2C in earnest once the care home evidence base is in place. Family referral from existing care home deployments seeds the channel from day one.

Three structural forces are colliding in the UK care sector, and none of them are reversible.

People are living longer.The fastest-growing age group is the over-85s — the cohort with the highest care dependency, the highest rates of dementia, and the greatest need for daily support. Their numbers will nearly double within a generation.

The population of older people is growing faster than the workforce that supports them. For every 1,000 people of working age today, there are 278 of pensionable age. By 2047 that rises to 302. The ratio is shifting structurally, not cyclically.

The people filling the gap are under severe and increasing pressure.Professional carers face chronic staffing shortages and rising documentation burdens. The July 2025 closure of the Health and Care Worker visa route removed the primary mechanism stabilising the workforce. Unpaid family carers — 5.8 million of them — absorb the shortfall at real cost to their own health, careers, and finances.

These forces are not future risks. They are present realities producing present consequences: undocumented care, missed deterioration signals, isolated residents, and families who cannot get a reliable picture of how their relative is doing.

England has approximately 17,000 care homes with around 460,000 registered beds for older people. The total UK adult social care market is worth £26.2bn annually, with residential and nursing care accounting for the largest share. Demand structurally outstrips supply: the 85+ population will nearly double to 3.3m by 2047, and projections suggest 200,000–470,000 additional care posts will be needed by 2040.

Supply is constrained on both ends. New construction is growing — contract awards reached £1.12bn in 2025, a 43% year-on-year increase — but planning delays, Biodiversity Net Gain requirements, and high build costs are limiting the pipeline. New builds cluster in affluent areas. Meanwhile legacy stock is closing: 21% of homes sold in 2025 were sold with vacant possession, and 40% of those left the sector entirely.

The result is a market with rising demand, constrained supply, and a widening quality gap between the homes absorbing investment and the homes falling behind. Both segments need what Iris provides, but for different reasons.

The care home market is not a single market. It is a bifurcated one, and the difference matters for every commercial decision that follows.

Care home ownership by sector

83% of care homes are private for-profit. 13% are voluntary or charity-operated. 4% are public sector. Nearly a third of all beds are owned by the 26 largest chains; the top five (HC-One, Four Seasons, Care UK, Barchester, Bupa) together provide around 50,000 beds, roughly 11% of the market. The remaining majority is independent, owner-operated, and small — most homes have 20–60 beds.

Private equity dominates the top of the market through sale-and-leaseback structures combined with operational debt loading. Two major collapses (Southern Cross in 2011, Four Seasons in 2019) together affected 45,000 residents. Independent operators watching that history take financial resilience and operational evidence seriously, because for them it is existential.

Care home residents are funded through three channels in roughly equal measure. Self-funders (45–50%) pay privately where their assets exceed £23,250, with average weekly costs of £1,298 residential and £1,535 nursing. Local authority-funded residents (40–45%) are funded by councils at approximately £951/week — £350/week less than self-funder rates at the same facilities. NHS Continuing Healthcare funds a smaller group (5–10%) at £1,500–£1,700/week.

How care home residents are funded

The gap between self-funder and LA rates means self-funders effectively cross-subsidise publicly funded residents. This produces a structural bifurcation in the sector:

Both segments are addressable. The pitch is different for each. This segmentation runs through the commercial and go-to-market strategy in Sections 6 and 7.

UK elderly population projections 2025–2047 (millions)

The 85+ cohort is the critical group. It is where frailty is most prevalent, where dementia rates are highest, where more than half have moderate or severe frailty, and where people are six times more likely to be admitted to hospital. By 2040, nearly one in seven people in the UK will be over 75.

The same cohort anchors both markets Iris addresses. Of the 1.7 million people aged 85+ today, only a minority are in care homes. The majority live at home — supported informally by family, largely invisible to any formal system. This is the B2C addressable population, addressed once B2B credibility is established.

Loneliness by age group (%)

Chronic loneliness in older people has been compared clinically to smoking 15 cigarettes a day. It is associated with a 40% higher risk of dementia, elevated cardiovascular risk, accelerated physical deterioration, and increased NHS admission rates.

The scale of loneliness inside care homes is particularly striking. Over 80% of nursing home residents report feeling lonely despite living in a community setting. The problem is not proximity. It is the absence of meaningful, personalised, ongoing connection. The same isolation exists, often more acutely, for older people living alone at home — where the absence of connection is total.

International recruitment vs. vacancy rate 2021/22–2024/25

The vacancy rate improvement between 2022 and 2025 was driven almost entirely by international recruitment. At its peak in 2023/24, an estimated 105,000 overseas workers started direct care roles. In July 2025, care workers and senior care workers were removed from the Health and Care Worker visa route. International recruitment has fallen by more than half since. The structural gap is re-opening.

Sitting on top of this shortage is the documentation burden. The average carer spends 40% of their shift on paperwork, and notes are written at the end of a 12-hour day from memory. Time recovered from documentation is the single most tangible value Iris delivers to the operator.

Impact of caring on unpaid carers' lives

Most unpaid carers are adult children, partners, or close family members of older people, caring at a distance, managing anxiety, and absorbing the weight of not knowing how their relative is doing day to day.

A phone call tells them how things are in that moment. A visit tells them what they can observe in an hour. The days in between are largely invisible. This is the emotional gap that Iris Connect closes — not through surveillance, but through the kind of warm daily update a trusted friend would provide.

For the B2C market, this group is the buyer. They are typically aged 45–65, digitally confident, balancing work and caring responsibilities, and already spending on peace-of-mind services — private GP access, home security, emergency alarms. What they cannot currently buy is a single product combining companionship, health monitoring, and family visibility at a consumer price point.

CQC inspections completed per year

Following the rollout of the CQC's Single Assessment Framework in 2023, inspection volumes collapsed. The backlog grows by approximately 424 locations per month.

The consequence: homes cannot rely on a timely re-inspection to recover a rating. They must demonstrate continuous, evidenced, person-centred care every day. This is the specific problem Iris solves for operators — not “better care notes,” but continuous regulatory evidence generated automatically as a byproduct of the product working. For a home waiting four years for re-inspection, that is the difference between demonstrating compliance and hoping for it.

The majority of the 85+ cohort live at home, not in care homes. The care home sector holds approximately 460,000 beds. The number of people aged 85+ living independently is substantially larger, and the gap widens every year.

The B2C addressable market is bounded by the number of older people living at home with families willing and able to subscribe on their behalf. That population is not only larger today, it grows faster — because the preference for independent living is strong and the care home sector cannot absorb demand at the rate it is arriving.

At £49 per month per household, the B2C TAM substantially outscales B2B. But this is a prize to unlock, not a market to launch into cold. The B2C purchase decision is triggered by a specific event: a fall, a health scare, a difficult visit, a phone call that did not go well. Families in that moment do not want an unproven product. They want something with a clinical track record, operator endorsement, and evidence that it works.

That track record is what B2B deployment produces. Care home adoption generates the clinical data, the case studies, the family exposure, and the word-of-mouth credibility that makes B2C credible at scale. B2B is the sequence. B2C is the prize. The full commercial logic of this sequencing is set out in Sections 6 and 7.

The pressures described here are not arriving sequentially. They are colliding simultaneously. The April 2025 NI increase and National Living Wage rise added £2.8bn to sector-wide costs. The visa closure removed the workforce safety valve. The CQC inspection backlog grows monthly. The 85+ curve is already underway. Family expectations have never been higher.

Technology adoption in care has historically been slow. The Digitising Social Care programme (£150m+ committed), the NHS 10 Year Plan's community care focus, and existential financial pressure on operators are changing that. The question is no longer whether care homes will adopt digital tools. It is which tools will earn enough trust to become the standard. The window to establish Iris as that standard is open now.

No UK product currently combines AI companionship, clinical signal generation, and CQC evidence output from a single system. That is the category gap Iris occupies.

Adjacent products exist. Some are companions without clinical infrastructure. Some are documentation platforms without the resident-facing layer that would generate the underlying observations. Some are monitoring tools that detect physical events but build no relationship with the person they observe. Each of these addresses part of the four-stakeholder problem — resident, carer, manager, family — but none address all four from one ecosystem.

This gap is structural, not accidental. Every existing product is architected around a primary buyer, and that buyer determines what the product optimises for. Documentation software is bought by managers and optimised for compliance. Family apps are sold as operator add-ons and inherit the constraints of the documentation platform beneath them. Companion devices are sold to families or procured through public health programmes and have no pathway into operational care infrastructure.

Iris is architected differently. The resident interaction is the primary data source, and every other stakeholder is served downstream of it. The carer's documentation, the family's daily update, the manager's compliance evidence, the clinical alert — all flow from the quality of a single resident conversation. This architecture is the reason one product can serve four stakeholders without compromise, and it is difficult to retrofit into existing products built around any other starting point.

ElliQ is the closest direct comparison to Iris. Built by Israeli company Intuition Robotics and launched commercially in the US from 2022, it is a purpose-built AI companion for older adults — a small illuminated robotic head with a companion touchscreen, designed to reduce loneliness and support ageing in place through proactive, personalised conversation.

The product has real strengths. It learns, remembers, initiates conversation, and has shown positive outcomes in small pilots. It has been deployed through US state ageing agencies in Florida, New York, New Jersey, and others, typically at no cost to the user, funded through public health contracts.

Despite four years of US commercial availability, ElliQ has not crossed into mainstream adoption. Distribution is concentrated in state agency procurement rather than direct consumer or operator sales, unit volumes remain modest, and there is no established UK commercial presence. For a UK operator or family evaluating options today, ElliQ does not exist as a practical choice.

Iris vs. ElliQ. ElliQ is a better-funded, more established companion product in the US public procurement channel. Iris is a more complete care system — built for the UK regulatory environment, designed for the operator as much as the resident, and present in the market where ElliQ is not.

Amazon launched Alexa Smart Properties for Senior Living in the UK in 2022, enabling care homes to deploy managed Echo devices. Alexa Together, the US family caregiver subscription, was discontinued in June 2024 and replaced by a more limited Emergency Assist product in May 2025.

Alexa is a general-purpose voice assistant adapted for care, not built for it. It has no knowledge of the resident beyond what is manually configured. It does not learn from conversation, does not build a longitudinal profile, and generates no clinical intelligence. Asking Alexa to play a song logs an activity; it does not capture the resident's mood, engagement, or the topics that brought them to life.

Alexa Smart Properties gives operators a managed voice assistant for announcements, smart home control, and resident-to-family calling. What it does not do: proactive companionship, personalised conversation, cognitive baseline tracking, care note generation, or CQC evidence documentation.

There is also a trust dimension specific to this context. Amazon is a US commercial platform whose business model is built on data. Operators procuring technology for vulnerable residents — and families choosing a companion for an elderly relative — are increasingly sensitive to data sovereignty and commercial intent. Iris is built on UK infrastructure, holds data in AWS London under UK data protection law, and has no advertising model or third-party data sharing. In a care context, that distinction matters.

Iris vs. Alexa. Alexa is infrastructure. Iris is a care product. The comparison is less a competitive threat than a demonstration of what a general-purpose assistant, built elsewhere for other purposes, cannot credibly become in the UK care environment.

The most commercially significant competitors for Iris are not the companion products. They are the UK care documentation incumbents whose relationships with operators Iris must work alongside or, ultimately, displace for the documentation workflow.

Nourish serves over 4,000 care providers across the UK and is the market leader in digital care planning. It offers a mobile app for carers to record observations at the point of care, alongside a family portal (Nourish Connect) and integration with NHS systems. Nourish has been investing in AI-assisted features, including suggested note content and anomaly detection.

Birdie is a well-funded domiciliary care platform that has expanded into residential, focused on real-time care recording, medication management, and family communication. It has raised substantial venture funding and is actively extending into documentation automation.

Person Centred Software (mCare) is the other major UK incumbent, deployed in several thousand homes, with a mature product covering care planning, medication, and audit.

These platforms share a fundamental architecture: they are tools that carers use to record care after it has been given. They rely on human input — a carer tapping observations into an app at the point of care, or writing notes at the end of a shift. They structure the evidence that humans provide; they do not generate the underlying observations.

Iris sits upstream of these platforms. It observes the resident directly and produces the substance of the care record automatically. A carer using both products would spend less time creating notes, not more — Iris drafts, the carer reviews, the documentation platform stores. Integration, not replacement, is the realistic near-term position.

Iris vs. documentation platforms. These platforms compete on the carer's time spent recording. Iris competes on whether the observation worth recording was ever generated at all. Over time, Iris's documentation output is likely to reduce the scope of what documentation platforms provide — but in the short term, coexistence is the more accurate frame.

Other UK documentation platforms in this space include CareLineLive, Log my Care, Radar Healthcare, KareInn, and Access Care Planning. The competitive logic is the same for all of them.

Products like Vayyar Care, Ally, personal alarm and fall detection devices, and acoustic monitoring systems address physical safety. Several are now present in UK care homes. They are passive by design — they watch for specific events and alert when something happens.

They do not interact with the resident, do not build a longitudinal profile, and do not detect the behavioural or conversational changes that precede physical events. Their output is an alert after the fact, not a pattern recognised in advance.

The clinical signal layer in Iris — cognitive baseline tracking, subtle engagement changes, mentions of physical discomfort — operates through conversational data rather than sensor data. The two categories are complementary in a home that uses both. They are not substitutes.

Family communication apps (Relatives Gateway from PCS, Nourish Connect, and care-home-specific portals) are typically add-ons to documentation platforms. Their content is only as good as the documentation beneath them. If the carer is busy, the update is late, thin, or missing.

Reminiscence therapy products — My Life TV, RemindMeCare, and similar — provide curated content for people with dementia. They are single-purpose activity tools, not companions or care systems, and generate no operational value for the home.

Iris generates family content directly from its observation of the resident, independently of carer workload, and integrates reminiscence naturally into conversation rather than as a separate activity session.

In the direct-to-family market, the competitive picture is thinner than in B2B. Existing consumer products for older people at home fall into two categories: emergency response (pendant alarms, fall detectors, GPS trackers) and logistics (medication reminders, digital GP access).

Both are reactive or task-based. Neither is relational. No product currently on the UK consumer market combines daily AI companionship, passive health monitoring, and a family visibility layer in a single subscription at a consumer price point. Iris is not competing with established consumer care technology. It is creating the category, once the B2B track record makes that category credible.

Rather than a tick-box matrix, the more useful comparison is along two axes: depth of operational integration (does the product sit inside the care home's daily operations, or alongside them?) and breadth of stakeholder coverage (how many of the four stakeholders does it genuinely serve?).

The most realistic competitive threat is not ElliQ entering the UK. It is Nourish, Birdie, or PCS adding a voice companion feature to an existing documentation platform.

This is a defensive move the incumbents will almost certainly make, and the architectural direction of travel across the sector is toward more AI-generated content. Iris's answer is not that incumbents cannot add voice. It is that adding a voice layer to a documentation-first architecture produces a different product than building a care system from the resident interaction outward.

A voice feature bolted on top of Nourish would sit above the existing documentation pipeline. It would not fundamentally change what data is captured, how the resident is known over time, or how the family experience is produced. It would be a feature; Iris is a system.

That said, incumbent response is a real risk and the time-to-value window matters. The case for moving quickly on UK pilot deployment and case study generation is not only commercial — it is defensive. Establishing the clinical track record and operator relationships before incumbents respond is the work of the next 12–18 months.

Iris's defensibility at this stage is not based on technology that competitors cannot replicate. It is based on four things that are hard to replicate quickly.

None of these are absolute defences. Together, they create a window — twelve to twenty-four months — in which Iris can establish the deployed base, the clinical evidence, and the operator relationships that make later competition a displacement problem rather than a category-creation problem.

Iris is a dedicated tablet, one per resident, that sits on the bedside table and moves with the resident around the home. It is not shared. Personalisation, memory, privacy, and continuous availability all depend on one device per person.

The interface is built around two natural ways of communicating: voice and touch. Say “Iris” and she responds. Or tap one of the large, clearly labelled options on screen. There are no hidden menus, no settings to get lost in, no steps that require remembering where you were. A resident who has never owned a tablet can use Iris. A resident with moderate dementia can use Iris.

Iris does not wait to be asked. She wishes the resident good morning. She checks in gently during the day. She lets them know when family have been in touch, when someone is calling, and when a favourite programme is about to start. She is present the way a good companion is — attentive, warm, and never intrusive.

From the moment Iris arrives, she already knows the resident. In care homes, staff enter the clinical and operational basics at admission — name, room, key preferences, anything already in the care plan. The family adds the personal layer through Iris Connect: photos, family names, stories, the grandchild's nickname, the dog she used to walk. Together, these two inputs produce a companion who introduces herself as someone who already knows the resident, rather than someone asking them to start from scratch. From there, every conversation enriches that picture naturally — Iris learns what the resident loves to talk about, what brings them to life, and what they would rather leave alone.

Iris also adapts to how someone speaks, mirroring the rhythm and warmth of the person she is talking to. For residents whose first language is not English, she can converse in the language they feel most at home in.

Family presence in the room.When a family member sends a photo through Iris Connect, Iris weaves it into conversation at a quiet moment: “Your daughter sent a lovely photo — she's been in the garden. Shall we have a look together?” When a grandchild records a voice message, Iris delivers it softly. When a family member calls, Iris announces it and the resident answers simply by saying yes. The family feels close. The resident feels remembered. No buttons, no confusion, no one feeling like a burden.

Entertainment and content.Television, radio, audiobooks, and music are integrated directly into the Iris experience. Familiar UK broadcasters and music services are available by voice or tap, without the resident navigating apps, menus, or logins. Iris integrates with major UK streaming and radio services, with credentials configured once by the family through Iris Connect — the resident never sees a login screen. Quizzes, word games, and reminiscence conversations are woven in naturally, drawn from the resident's own life and interests rather than from a generic template.

For residents who find comfort in knowing what the day holds, Iris offers gentle orientation — what day it is, when meals are, whether family have been in touch. The daily plan, set by the care team in Iris Care, is always there if the resident wants to know what is coming next.

The device as a service.Iris owns and maintains every device. It arrives pre-configured. If it ever needs replacing, a new one ships within 48 hours, already set up with everything Iris knows about the resident. At the end of a subscription — or when a resident passes away — the device is reset, the resident's data is cleared in line with retention policy (Section 5), and the device is either reassigned to a new resident or returned to Iris. There is nothing for the home or family to procure, manage, or return.

Iris Connect is for the people who love an elderly relative, whether they are living at home or in a care home. It is a way to stay close, share moments, and feel part of their daily life — without requiring anything of the resident.

Connection.Families can call or video call at any time; Iris announces the call on the device and the resident answers by voice. They can send photos, which Iris weaves warmly into conversation rather than delivering as a notification. They can record a voice note — a few seconds from a grandchild, a hello from a son — and Iris finds the right moment to play it.

A daily sense of how things are going.Families receive a warm daily note: not a clinical summary, but the kind of update a trusted friend would give if they had popped in. How the day started. What brought her to life today. Whether the family's photo was received warmly.

Setup from a distance.Families build the resident's world through the app, doing the configuration work once so their relative never has to think about it. Streaming services and preferred content can be set up remotely — the family logs in on their relative's behalf through Iris Connect, and the content simply appears, ready to use, with no passwords for the resident to remember and no technical steps for anyone to walk them through over the phone. Browser bookmarks work the same way — a local newspaper, a grandchild's school page, a familiar shop — all reachable by voice or tap, configured by the family, invisible to the resident.

Everything the family shares through Iris Connect becomes part of how Iris knows and talks to their relative. A name added to the contacts list means Iris introduces callers correctly: “your grandson Leo is calling” rather than an unknown number. A photo uploaded to the memory collection becomes a conversation starter. A preference noted in the app shapes how Iris engages day to day.

For families whose relative is living at home (B2C), Iris Connect is the whole picture — daily connection, setup, and the quiet reassurance that someone is there. For families whose relative is in a care home (B2B), Iris Connect sits alongside Iris Care, giving them a direct line to their relative that does not depend on the care team having time to write an update. Where the care home has enabled it, families can also see approved care notes in a read-only view — but this is off by default and always at the discretion of the care team.

Iris Care serves two people simultaneously: the carer on the floor and the manager responsible for the home. The two views share a data substrate but present different interfaces, because the two roles need different things.

The average carer spends 40% of their shift on documentation, writing notes from memory at the end of a 12-hour day. Iris Care replaces that. Everything Iris has observed through the day — conversations, medication confirmations, mood, engagement — is compiled into a structured draft care note. The carer opens it, adds their own physical observations, and confirms it with a single deliberate action: their professional sign-off. Notes that previously took 45 minutes take 90 seconds. And because they are built from real observation rather than end-of-shift recollection, they are more complete, more consistent, and more defensible than anything written from memory.

The time recovered is time returned to care. When documentation takes care of itself, carers can be where they are needed — with the people in their care, not at a desk with a keyboard.

The carer view also provides a home-wide operational picture. A colour-coded grid shows at a glance which residents are settled and which might benefit from a visit, without waiting for someone to raise a concern or press a call button. Tapping any resident brings up the detail: how their day has gone, what has been noted, whether family have been in touch. At shift change, a structured handover covers the whole home — overnight events, outstanding items, anything needing follow-up — so nothing falls through the gap between teams.

Clinical prompts are present but deliberately understated. Subtle changes in how a resident is engaging — quieter than usual, a shift in response patterns — are surfaced gently as a prompt to follow up. Not diagnoses. Not demands for immediate action. A nudge from a system that has been paying close attention, so the carer can make an informed decision.

For the manager, Iris Care addresses three of the defining pressures of the job: CQC inspection risk, staff retention, and operational visibility.

On compliance, every interaction is logged, timestamped, and attributed. The result is continuous, audit-ready evidence generated as a natural byproduct of Iris working — not assembled separately under pressure. When an inspector asks for evidence of person-centred care, the manager exports a complete, formatted record for any date range. It was being generated every day. A home that uses Iris does not prepare for inspection; it is always ready.

On staff retention, the picture is quieter but just as significant. One of the most cited reasons carers leave is the burden of administration that crowds out the reason most of them came into the work in the first place. A home that can offer a more modern, less administratively burdened way of working is better placed to attract and hold on to good people in a market where good people are hard to find.

On operational visibility, the manager view provides aggregated analytics across the resident population — engagement trends, documentation completion rates, alert response times, family engagement metrics — replacing the patchwork of memory, paper, and end-of-shift guesswork that many homes still rely on.

Iris Care is not just an efficiency tool. It is the infrastructure for a better-run home — one where the manager has the information they need, the staff have the support they need, and the residents receive the care they deserve.

Information in the Iris ecosystem travels in three distinct directions, and those directions do not cross.

The care team receives what they need to provide good clinical care: medication confirmations, safety signals, AI-drafted care notes, and subtle changes in engagement that may warrant a follow-up. The family receives what keeps them connected and reassured: a warm daily note, confirmation that their messages and photos were received. The elderly relative has a private space, things said in confidence that belong to them alone, never surfaced to carers or family unless there is a genuine concern for her physical safety.

The level of information shared with the care team and family is set at onboarding and can be adjusted, but the private space is inviolable at every level. It cannot be unlocked by a manager, a family member, or a carer. It exists because privacy and dignity do not stop being important when someone moves into a care home or needs more support at home.

Iris follows industry best practice across every aspect of data security and privacy. All data is encrypted both in transit and at rest, with regular key rotation as standard. All data is stored in the AWS London region, ensuring it never leaves the UK and remains fully subject to UK data protection law. No raw audio is ever retained, conversations are processed in real time, and only structured summaries and relevant care signals are stored. Full conversation transcripts are not kept.

Consent is recorded at onboarding with a timestamp. Where a resident has full capacity, consent is their own. Where capacity is reduced, consent is obtained from the holder of Power of Attorney. Every resident has the right to have their profile and all associated personal data deleted on request. Care notes are retained for seven years as required by CQC, but personal profile data, conversation summaries, and family connection data are removed in full.

Data Processing Agreements are in place with all third-party processors. A Data Protection Impact Assessment is available on request.

Iris does not retain raw audio at any point or for any purpose. It does not store full conversation transcripts — only structured summaries and event flags derived from them. It does not record or store carer voice, actions, or identity during care interactions. And no data from one person's account is ever accessible from another — every resident profile is fully segregated, with access controls that make cross-account access structurally impossible rather than merely prohibited by policy.

Iris is designed to meet the requirements of UK GDPR, the Data Protection Act 2018, and the CQC's expectations around data governance in registered care settings. Data residency in AWS eu-west-2 (London) means all personal data remains within the UK. The append-only audit trail — covering every care note, every alert acknowledgement, every consent record — is structured to withstand regulatory scrutiny and supports the CQC evidence framework described in the Iris Care section.

For care home operators, Iris operates as a Data Processor under a formal Data Processing Agreement. The care home remains the Data Controller for resident information. For families subscribing directly through Iris Connect, Iris acts as both Data Controller and Processor for personal data held within the platform.

Iris serves two markets in deliberate sequence: B2B as the proving ground, B2C as the expansion. Within each, there are distinct buyer archetypes, and the commercial approach is tailored to each.

Three buyer types, each with a different trigger, sales cycle, and conversion path.

One buyer archetype, one trigger mode.

The document has made the case throughout that B2B comes first not because it is the larger market, but because it is the credibility substrate that makes B2C viable at scale.

The sequenced approach runs as follows.

This sequencing shapes everything that follows in Sections 7 and 8.

The Iris device is included in every tier — pre-configured, shipped, maintained, and replaced. There is no hardware procurement, no IT setup, no capital expenditure for the subscriber. Every subscriber gets the full product; there is no stripped-down version and no upsell conversation.

The consumer purchase is emotionally motivated and the value is immediately legible. At £49/month, Iris sits within the range families already spend on peace-of-mind services: private GP access, home security, emergency alarm subscriptions. One hour of private home care costs £25–35. For less than two hours of care per month, a family gets daily companionship, ongoing health monitoring, family visibility, and the reassurance that someone is always there.

The purchase is not being compared to other AI companions; it is being compared to the alternative of doing nothing, and to the guilt that accompanies that alternative.

B2B churn is low by structure — 24-month contracts, high switching costs, operational integration. Assumed annual churn: 8%.

B2C churn is low by relationship — once Iris is part of a household's daily routine, once the elderly relative is having conversations with her, the activation energy required to cancel is high. Assumed annual churn: 15%.

B2B contracts are 24 months standard, billed monthly per active bed. The per-bed model means subscription revenue scales naturally with occupancy — a home running at 85% occupancy pays for 85% of beds, not the full 40.

Group contracts extend to 24 months at £25/bed with a minimum commitment of 5 homes. Breaking the contract requires written notice at the renewal point; early termination is not permitted outside specified conditions.

B2C subscriptions are 12 months, billed monthly, with a 30-day free trial. Churn in this model is tracked monthly but measured annually for cohort analysis.

Expansion revenue in B2B comes from occupancy growth (existing homes filling previously empty beds) and from operator expansion (a single-home customer adding their second home, a group director rolling out to additional sites). This is the expansion engine we expect to dominate B2B growth from Year 2 onward.

Expansion in B2C is primarily referral-driven — a subscribing family introduces a sibling, a colleague, or a friend in a similar situation. The referral programme (discussed in Section 7) is structured to make this easy and to reward it meaningfully.

The goal of Phase 1 is not revenue. It is evidence that will support everything that follows.

Three to five care homes, selected as design partners. The selection should span LA-heavy and self-funder-heavy segments, small independents and small groups, to ensure the case studies that emerge are credible across the full range of operator conversations. Geographic clustering matters — regional concentration creates word-of-mouth within local networks, care association meetings, and sector press rather than isolated proof points.

Every pilot runs on a free 30-day basis and converts to a paid 24-month contract at the end. The commercial target for Phase 1 is not pilot count; it is conversion rate and evidence quality.

The output of Phase 1 is a case study pack: quantified carer time saved, CQC documentation generated as a byproduct, family engagement rates, resident interaction data, and at least two care home managers willing to speak on record. That pack is the primary sales tool for every conversation in Phase 2 and beyond. Nothing sells Iris to a care home manager more effectively than another care home manager who has run the pilot and signed the contract.

Phase 2 opens outbound sales in B2B while quietly launching B2C to a hand-picked cohort.

By Month 12, Iris has two to three published case studies from Phase 1 pilots, a validated B2C cohort, and enough clinical data to stand behind the product in public marketing. Phase 3 is where the consumer category creation work begins in earnest.

The consumer launch opens three acquisition channels simultaneously.

The B2B motion accelerates. A second sales hire joins at Month 15 to expand coverage beyond the initial regions. The Care Show presence upgrades to exhibiting from Year 2 — a £20–30k stand investment that is no longer speculative because pilot case studies give the stand something real to show.

Group conversations begin in earnest. Most UK care groups will not engage seriously until there are 15–20 live homes as reference points. By Month 18, Iris is in that range and can credibly enter procurement cycles with groups of 5–50 homes. These are longer cycles (6–12 months), but the deal sizes justify the wait.

Cross-channel reinforcement. Every care home deployed in Phase 3 continues to expose families to the product. A prompt at resident admission invites families to subscribe directly through Iris Connect if they have another relative living at home. This is the highest-converting B2C lead source and costs nothing beyond the onboarding workflow.

By Month 24, Iris is running two motions simultaneously. B2B is the recurring revenue backbone and the ongoing credibility machine. B2C is the growth story, with CAC held down by family referral, care-home-derived leads, and the content library that now ranks for trigger-moment search.

Phase 4 is when the compounding dynamics show up. The installed B2B base generates B2C leads continuously. The B2C subscriber base generates referrals that feed both consumer and care home conversations. The published case studies accumulate. Sector presence becomes established. Within this, the B2B motion moves upmarket into group deals and the earliest ICS commissioner conversations, while B2C enters mainstream category awareness.

This is the point at which Iris starts to look less like a care tech startup and more like a UK consumer ageing brand. That is the destination.

The primary B2B channels are sector-specific. These are credibility channels as much as marketing ones — presence in them signals that Iris belongs in the sector conversation.

PR strategy is care-sector-specific rather than aimed at general technology press. A feature in CMM Online reaches the decision-maker that TechCrunch does not. Fractional PR support starts at Month 9, scaling with deployment volume.

The B2C acquisition strategy is built around three mechanisms that reinforce each other over time.

Four partnership categories, pursued in sequence.

UK first, proven, then international.

The US market is the largest single international opportunity, but it requires regulatory re-engineering (HIPAA, state-level variation), a local commercial presence, and a distribution model that does not depend on UK care sector relationships. Realistic US entry is a Year 3+ conversation, built on a dominant UK position rather than in parallel with UK execution.

Europe — specifically Germany and the Netherlands, the markets most structurally similar to the UK — becomes accessible once a regional commercial presence is established. These are also Year 3+ considerations.

International sequencing is deliberate. Spreading too early dilutes the evidence base and the team. A dominant UK position, with the case studies and clinical data to prove it, is the foundation that makes every international conversation possible.

The model runs over three years, quarterly, with annual summaries for this document. It reflects the sequencing set out in Section 6: B2B pilots from Month 0, quiet B2C from Month 3, public B2C from Month 9–12.

Founder and team. The founder is the sole technical resource and takes no salary across the three-year projection. This is a founder subsidy, not a market-rate structure; investor readers should impute a £100k/year technical founder cost when evaluating the business on a normalised basis. First commercial hire: BDR / sales at Month 6 (£55k base + £15k OTE). Second commercial hire: Customer Success at Month 9 (£45k). Third hire: second salesperson at Month 15 (£60k + £20k OTE). Additional CS hire at Month 24 as deployed base passes ~100 homes.

Pricing. Flat as set out in Section 6.3: £30/bed/month standard B2B, £25/bed/month group (20% of homes), £49/month B2C.

Growth targets. 40 homes live end of Year 1, 130 end of Year 2, 320 end of Year 3. B2C subscribers: 200 end of Year 1, 800 end of Year 2, 2,500 end of Year 3.

Churn. B2B 8% annual. B2C 15% annual.

AI token cost. Token input cost halves annually — a conservative assumption against recent market trend.

ARR is the exit-run-rate figure at year-end. Recognised revenue in-year is lower because homes and subscribers ramp through the year rather than being live from Month 0.

Total ARR & gross profit (5-year trajectory)

Marketing spend scales with B2C public launch in Year 2 and full dual-motion in Year 3. The Year 3 figure assumes paid B2C CAC of ~£250 and B2B sector marketing at sustainable levels, not aggressive growth.

At 40 homes × 40 beds = 1,600 devices by end of Year 1, plus 200 B2C devices = 1,800 devices. At £80/device, that is £144,000 in hardware to deploy. Because OEM lead times are 8–12 weeks with MOQ constraints, hardware is purchased in batches ahead of deployment.

The hardware is amortised over 36months for P&L purposes, but the cash outflow is front-loaded against deployment.

Annual operating cash flow vs. cumulative position

On these numbers, the minimum cash requirement to reach cash-flow positive is approximately £250k of working capital, assuming founder salary remains zero. Adding a market-rate founder salary (£100k/year) across the three-year period adds £300k to the requirement, bringing the realistic funding ask to £550–750k for a lean-growth scenario. A faster-growth scenario — accelerated sales hiring, aggressive B2C paid acquisition, geographic expansion — would push the ask to £1.5–2.5m.

This is an output of the model rather than a stated funding ask. The appropriate round and amount depend on the pace of execution the business wants to underwrite and the milestone the funding needs to reach.

B2B global sub-total: ~£2.8bn annually.

B2C global sub-total: ~£19.7bn annually.

B2C TAM is calculated on the assumption that every household with an 85+ relative is addressable; a more conservative figure filtering for household income and digital confidence would reduce the UK TAM to ~£350–500m. The larger figure reflects total addressable population; the conservative figure reflects realistically marketable population. Both are useful framings for different audiences.

This represents 2.8% of UK B2B TAM and 0.18% of UK B2C TAM.

Because B2B contracts are 24-month commitments, reported ARR understates committed revenue. At end of Year 3, the contracted backlog (total committed revenue from existing B2B contracts beyond the reporting date) is approximately £8,908,800, producing a Total Contract Value (TCV) of £14,833,200 at end of Year 3.

This provides a guaranteed revenue floor that differentiates Iris from a pure consumer subscription business and supports stronger valuation multiples at future funding rounds.