For families
Understand early changes with clarity, and know when it may help to seek support.
Learn more
Responsible, evidence-led, human-first AI
AI-powered insight for earlier dementia detection
DementiaDetect is being developed to identify early signals of cognitive decline, helping people, families and clinicians move from uncertainty to earlier understanding, support and action.
Designed to support, not replace, clinical judgement.
Speech · Sleep · Steps · Signals
Everyday data, brought together with care
Dementia does not usually appear in a single moment. It often emerges through small changes over time. DementiaDetect is designed to listen for those changes across multiple signals.
Speech
Changes in language, fluency and word-finding.
Sleep
Shifts in rest patterns and overall brain health.
Steps
Movement, activity and changes in daily routine.
Signals
Brought together with AI to surface patterns.
The problem
Dementia is too often detected late
For many people and families, the first signs of cognitive change are subtle. A shift in speech. A change in sleep. A difference in routine. A moment of confusion that feels easy to explain away.
Too often, these signals are only acted on once the impact has become harder to ignore. For families, that can mean months or years of uncertainty. For clinicians, subtle change can be difficult to assess without clear, consistent information over time.
DementiaDetect exists to help change that.
~1m
people are estimated to be living with dementia in the UK today
1.4m
projected by 2040 as the population ages
1st
the NHS highlights the benefits of accurate and earlier diagnosis
Sources: Alzheimer's Society and NHS England. Figures are estimates and projections.
The approach
Listening to the signals that matter
DementiaDetect uses AI to analyse patterns across multiple brain-health indicators, including areas such as speech, sleep, movement, cognition and behaviour, to support earlier awareness and clearer next steps.
Capture everyday signals
With consent, DementiaDetect brings together relevant brain-health and behavioural information from digital tools, assessments and connected devices.
Detect meaningful patterns
AI models look at how signals change over time, helping surface patterns that may be associated with cognitive change.
Support earlier action
Insight can help people, families and clinicians decide when further assessment, monitoring or support may be helpful.
What it could look like
From signals to a clear, gentle summary
We are designing DementiaDetect to turn everyday signals into something calm and understandable: a simple view of how things are changing over time, and a clear sense of when it may help to speak to a healthcare professional.
- A simple brain-health summary, not a score to fear
- Change over time across speech, sleep and activity
- Plain-language, non-diagnostic insight
- Clear guidance on sensible next steps
- A summary you could choose to share with a clinician
Brain-health summary
Illustrative
Speech
Slight change
Sleep
Stable
Activity
Stable
Change over time
Last 12 weeksExample insight. A small change in language patterns over recent weeks. It may help to discuss this with a healthcare professional.
Illustrative preview only. Not a diagnosis and not a real result. Final design and outputs are being developed with clinical input.
Why it matters
Earlier insight gives people something dementia too often takes away: time.
Earlier insight cannot remove the difficulty of dementia. But it can give people and families more time to understand, plan and access support.
- Time to seek advice
- Time to involve family
- Time to plan care
- Time to access support
- Time to make informed choices
- Time to live with greater clarity
Why now
The moment for earlier detection is here
Need, data and technology are converging. That is why we are building DementiaDetect now, and building it responsibly.
Rising prevalence
Around 1 million people in the UK live with dementia today, projected to reach 1.4 million by 2040.
Pathways under pressure
Memory services and diagnosis routes are stretched, and recognition often comes late.
New everyday data
Smartphones and wearables now create passive signals that did not exist a decade ago.
AI that reads patterns
Modern AI can analyse subtle, longitudinal change across many signals at once.
Systems need earlier models
Health systems increasingly need earlier, scalable and responsible approaches to care.
The cost of waiting
Every late recognition is lost time for families, planning and support.
Built on responsible AI
AI for brain health must be built with care
Dementia detection is sensitive and serious. DementiaDetect is being developed with a commitment to privacy, transparency, fairness, clinical validation and human oversight from the beginning.
Privacy by design
Personal health data is handled with high standards of security, consent and transparency.
Human oversight
Our tools are designed to support healthcare professionals, never to replace them.
Fairness and evidence
Models are developed to be tested for bias, and claims are kept proportionate to validation.
Where we draw the line
What we do not claim
Being honest about limits is part of being trustworthy. So we are clear and consistent about what DementiaDetect is not.
We do not diagnose dementiaAny insight is a non-diagnostic signal, never a medical conclusion.
We do not replace cliniciansWe are designed to support clinical judgement, not stand in for it.
We do not promise certaintyWe surface patterns that may warrant attention, not guarantees.
We do not sell fearOur aim is calm, dignity and earlier support, not alarm.
Who it is for
One mission, many partners
For clinicians
Support earlier risk insight, longitudinal monitoring and assessment pathways.
Explore clinical useFor partners
Collaborate on validation, pilots and responsible AI development.
Partner with us
Our story
It began with my Grandma
DementiaDetect began with something deeply personal.
Like so many families, we did not realise at first that the small changes we were seeing might mean something more. A moment of confusion. A repeated question. A word that would not quite come. Individually, easy to excuse. Together, they were trying to tell us something. But we did not know what we were seeing, and that is the part that stayed with me.
I cannot change what happened in my own family. But I can help build something that gives other families more clarity, more confidence and more time.
Help shape the future of earlier dementia detection
Whether you are a clinician, researcher, health system partner, investor or someone with lived experience, we would welcome your interest.