Healthcare doesn’t lack compassion, intelligence, or effort. What it often lacks is a reliable way to connect the right people, the right data, and the right questions—fast enough to matter.

Across cancers, motor neurone disease (MND), liver failure, kidney failure, and many other serious conditions, the same pattern repeats:

• Patients and carers live the reality every day, but their lived experience is rarely captured in a consistent, research-ready way.
• Clinicians hold essential insights, but those insights are scattered across notes, systems, and sites.
• Researchers move carefully (as they should), but the time between “signal” and “proof” can stretch into years.
• Foundations and research centres do extraordinary work, yet often have to rebuild infrastructure for every new program, registry, or study.

The result is a frustrating mismatch between urgency and pace.

This blog is a public invitation from Digital One Agency to health foundations and research centres worldwide: let’s form joint ventures (JVs) that build ethical, privacy-first AI and data infrastructure—shared rails that make research and real-world learning faster, more inclusive, and more useful for everyone.

Not as a “vendor relationship”. Not as a one-off pilot. And not as a promise that technology alone can solve complex disease.

Instead, as a collaboration where:

• foundations and research centres lead the mission and governance
• patients and clinicians shape what’s captured and why
• and Digital One Agency contributes engineering, data systems, and applied AI capability to help transform real-world information into real-world progress.

What follows is the “why”, the “how”, and the practical shape of what we can build together—starting with cancer and MND, while also supporting other high-impact health areas.

Why This Matters Now (and Why It’s Bigger Than Any One Organisation)

For many conditions, we already have promising therapies, supportive interventions, and care models. But outcomes can still vary dramatically by:

• disease subtype and biology
• symptom patterns and progression pathways
• comorbidities
• geography and access to care
• adherence and tolerability
• timing (how early someone is diagnosed, referred, or supported)
• social determinants (housing, transport, finances, mental health support)

When we don’t measure these factors consistently, we can’t learn from them at scale. We end up with:

• small datasets that struggle to reflect diversity
• fragmented evidence across sites and countries
• delays in identifying what actually changes outcomes
• an over-reliance on “averages” that don’t help individuals

AI can help—but only if it’s built on strong foundations: consistent data collection, careful governance, and shared ownership of the mission.

That’s where foundations and research centres are uniquely powerful.

Why Foundations and Research Centres Are Ideal Leaders of Ethical AI Collaboration

Foundations bring trust and community accountability

Many patients are understandably cautious about data, privacy, and commercial motives. Foundations—by virtue of their mission and accountability—are often best placed to set expectations, communicate clearly, and maintain trust.

Research centres bring rigour and clinical validation

AI should not be “deployed” into health contexts without robust evaluation. Research centres can define appropriate methods, governance, and validation pathways.

Together, you can make data useful without making it extractive

The goal isn’t to “collect more data”. It’s to create a system where:

• patients understand what’s happening
• consent is meaningful and changeable
• data use is transparent and auditable
• research access is managed responsibly
• outputs are validated and clinically relevant

A JV structure can support shared stewardship and long-term sustainability, rather than short-term projects that fade when funding cycles end.

What We Mean by “AI Solutions” (and What We Don’t)

When people hear “AI in health”, they often picture chatbots or black-box predictions.

That’s not the centre of this invitation.

The most valuable AI in healthcare is often quiet and practical. It helps by:

• structuring messy information (turning notes into consistent research fields without losing nuance)
• finding patterns across time (not just single appointments)
• identifying subgroups (patients who share symptom clusters, trajectories, or therapy responses)
• supporting earlier intervention (flagging risk trends that clinicians can review)
• speeding up research workflows (cohort discovery, trial matching, outcome tracking)
• closing the loop (turning real-world outcomes into improved guidelines and support programs)

What AI should not do:

• replace clinicians
• make unreviewed clinical decisions
• “predict” without transparency and validation
• compromise patient privacy or consent integrity
• prioritise growth over safety and trust

In our view, AI earns a place in healthcare only when it is explainable, governed, measured, and helpful—and when people remain in control.

The Core Concept: Continuous, Longitudinal Learning (Not One-Off Data Collection)

Many registries and studies are built for a narrow purpose, then retired or underused. But serious conditions are not one-off events; they are journeys—often spanning years.

A more effective approach is to build a continuous learning system that can support:

• registries
• observational studies
• supportive care programs
• trial recruitment and follow-up
• outcomes tracking
• research collaborations across sites and borders

What makes continuous learning different?

• Data is updated over time (symptoms, interventions, outcomes)
• Patient-reported outcomes are included meaningfully
• Clinician input is captured in consistent formats
• The platform becomes a living resource for approved research
• Insights improve programs, not just publications

The practical benefit: each new participant and follow-up strengthens the evidence base, rather than restarting the clock each time.

A Joint Venture Approach: Shared Stewardship, Shared Capability, Shared Benefit

“JV” can mean different things in different contexts. In this invitation, a JV is less about corporate structure and more about shared stewardship:

• Shared governance: foundations and research centres help define what data is collected, how it’s used, and what “success” looks like.
• Shared build: Digital One Agency contributes engineering and AI implementation; partners contribute domain leadership, community engagement, and research direction.
• Shared accountability: clear ethics processes, oversight, and transparent reporting.
• Shared sustainability: a platform that can be maintained, improved, and scaled without rebuilding from scratch.

Importantly, a JV can be designed to align with mission-led values:

• patient interests protected first
• consent and privacy treated as core design features
• research access controlled, audited, and purpose-limited
• long-term outcomes prioritised over short-term outputs

There is no one-size-fits-all model here. Different conditions, regions, and partner ecosystems require different governance and operating structures. The key is building something that lasts and learns.

The Platform We Can Build Together: The “Research Rails” That Many Conditions Can Share

Below is a practical blueprint for an ethical, foundation- and research centre-led AI research platform. It can start small and grow—without losing integrity.

1) Patient registry and consent management that respects choice

A registry is not just a database; it’s an agreement with people.

Key features:

• clear consent pathways (what is collected, why, who can access, for what purpose)
• options to participate at different levels (e.g., clinical data only, PROs, wearables, imaging, genomics)
• ability to withdraw or change consent over time
• plain-language explanations and ongoing communication

2) Patient-reported outcomes (PROs) designed for real life

PROs are often underpowered because they’re cumbersome. Done well, they can be transformative.

Examples of what can be captured:

• symptom severity and variability
• fatigue, pain, sleep quality
• functional status (mobility, swallowing, speech, daily tasks)
• mental health and wellbeing indicators
• treatment side effects and tolerability
• “what happened this week?” narrative capture
• carer observations (where appropriate)

3) Clinician and specialist workflows that reduce burden

Clinicians are time-poor. A system that adds friction won’t scale.

Approach:

• capture a small set of validated clinical measures
• align with existing workflows where possible
• allow structured fields plus short narrative
• use AI to assist with structuring notes (with clinician review)
• minimise duplicate entry by integrating with existing systems when feasible

4) Integration with medical data sources (as available and consented)

Depending on jurisdiction and partner access, the platform can integrate:

• pathology and lab results
• imaging reports and metadata
• medication history and dosage changes
• hospital admissions and discharge summaries
• oncology treatment protocols and cycles
• renal dialysis parameters
• respiratory function tests
• genetic testing results (where appropriate and consented)

Not every partner will start with integrations. Many programs can begin with a registry + PROs + clinician input, and add integrations progressively.

5) Data standardisation that enables collaboration across sites and borders

If every site calls the same concept by a different name, research slows.

Where appropriate, we align with widely used healthcare data standards and vocabularies (while keeping the system usable for non-technical users). The aim is not perfection—it’s consistency good enough to support valid analysis.

6) Privacy, security, and auditability built in, not bolted on

Trust depends on design choices:

• role-based access controls
• full audit logs of data access and changes
• de-identification and pseudonymisation strategies
• secure environments for approved researchers
• data minimisation (collect what’s needed; avoid what isn’t)
• clear processes for ethics approvals and data use agreements

7) Research workspaces that make approved analysis easier

Instead of exporting spreadsheets over and over, a platform can provide:

• controlled research datasets for approved studies
• cohort-building tools (find eligible participants ethically)
• outcome tracking dashboards
• reproducible analysis pipelines
• documentation and data dictionaries

This reduces the repeated overhead that slows research centres and makes partnerships harder.

What AI Adds on Top: Practical Capabilities That Improve Over Time

Once the “rails” are in place, AI can support meaningful workstreams.

A) Pattern detection and subgroup discovery

Many conditions are not one disease but multiple subtypes under one name.

AI can help identify:

• symptom clusters
• progression pathways
• response subgroups (who benefits from what)
• early warning trends that deserve clinical attention
• care pathway differences correlated with better outcomes

B) Natural language processing (NLP) to unlock clinical narrative

So much value is trapped in:

• referral letters
• specialist notes
• discharge summaries
• MDT meeting notes
• patient narratives

NLP can help structure this information into research-ready variables—while preserving original text for context and audit.

C) Therapy outcome tracking in the real world

Clinical trials are essential, but real-world settings introduce complexity:

• comorbidities
• polypharmacy
• adherence challenges
• different dosing and supportive care strategies
• varied access to allied health

AI can help analyse what’s happening in practice, with careful methods and appropriate caveats.

D) Trial and study matching support

Many people who could participate in research never hear about opportunities.

With proper consent and governance, the platform can support:

• identifying eligible cohorts
• notifying sites or study coordinators
• tracking recruitment and retention barriers
• improving equity in participation

E) Continuous improvement loops for support programs

Foundations often run support services—education, navigation, peer support, financial assistance, equipment support.

A platform can help evaluate:

• what services are used
• when people access them
• what outcomes are associated with timely support
• where gaps persist

This is not about “optimising” people. It’s about ensuring help arrives earlier and more effectively.

Condition Areas We Can Build For (Starting With Cancer and MND)

This invitation is open to all health foundations and research centres, but we’re explicitly focusing on cancers, motor neurone disease, and other serious conditions where longitudinal learning can improve outcomes and quality of life.

Below are examples of condition areas where this approach can be immediately relevant.

1) Cancer: From fragmented journeys to coordinated learning

Cancer research is advanced, and yet many challenges remain:

• treatment selection and sequencing complexities
• side effect management
• survivorship and late effects
• quality of life trade-offs
• rare cancers and small cohorts
• inequity of access to trials and specialist care

A foundation- and research centre-led AI platform can support:

Longitudinal patient journey capture

• diagnosis, staging, biomarkers (as available)
• treatment protocols, cycles, and modifications
• side effects and symptom burden across time
• supportive care interventions and outcomes
• hospitalisations and complications
• survivorship tracking and late effects

Subgroup insights

• identifying symptom and side-effect phenotypes
• correlating supportive care strategies with adherence and wellbeing
• detecting patterns that may suggest earlier intervention needs

Trial readiness and recruitment

• ethically supported cohort discovery
• improved diversity in research participation
• retention insights (why people drop out and how to reduce burden)

Cancer types where foundations and research centres often seek better “real-world learning”

• breast, colorectal, lung, prostate, ovarian
• melanoma and other skin cancers
• pancreatic and liver cancers
• paediatric cancers
• haematological cancers (leukaemia, lymphoma, myeloma)
• rare cancers where small numbers make learning hard

This approach isn’t about replacing trials. It’s about creating an evidence layer that complements trials and strengthens translation into practice.

2) Motor Neurone Disease (MND): Capturing progression, function, and lived experience

MND is a condition where time matters enormously—both medically and practically. Support, equipment, allied health, communication tools, and respiratory planning can change quality of life.

A strong, continuous learning system for MND can help capture:

• onset patterns (bulbar, limb, mixed presentations)
• functional scoring over time
• speech and swallowing changes
• respiratory function indicators (as available)
• nutrition and weight trends
• assistive technology needs and timing
• carer burden and support needs
• service access and delays (where the system is failing people)

AI can help identify:

• progression trajectory clusters (without pretending every individual is the same)
• early signals that correlate with faster decline
• supportive interventions that correlate with better quality-of-life outcomes
• care pathway patterns that reduce crisis events and unplanned admissions

Just as importantly, an MND-focused platform can provide foundations and research centres with:

• a research-ready cohort built on consent and transparency
• consistent follow-up structures
• a way to unify specialist and community data respectfully

3) Liver failure and severe liver disease: Seeing decompensation earlier, learning from longitudinal signals

Liver failure and advanced liver disease often involve complex, fluctuating trajectories—periods of stability punctuated by sudden deterioration.

A shared platform can support:

• longitudinal labs (where integrated)
• symptom patterns, fatigue, cognition changes, sleep, appetite
• medication changes, adherence barriers, side effects
• complications and admissions
• transplant pathway milestones (as appropriate)

AI can help:

• identify patterns that precede decompensation events
• surface care pathway differences that correlate with fewer admissions
• understand which supportive strategies improve functional outcomes and wellbeing

4) Kidney failure and chronic kidney disease: Improving pathways, not just measuring outcomes

Kidney disease is often a long journey. Dialysis, transplantation, cardiovascular comorbidity, and medication burden create layers of complexity.

A foundation- and research centre-led AI system can support:

• CKD progression tracking
• dialysis parameters and wellbeing outcomes (where applicable)
• medication profiles and side effects
• mental health, fatigue, sleep, appetite and functional capacity tracking
• service access patterns (transport, scheduling burdens, regional inequities)

AI can help:

• identify subgroups with faster progression risk signals
• understand the impact of supportive services on retention and quality of life
• support earlier referral and planning (with clinician oversight)

5) Other high-impact areas foundations frequently lead

While our starting focus is cancer and MND, the same infrastructure can support multiple serious health domains. Examples include:

Cardiovascular

• heart failure, cardiomyopathies
• arrhythmia monitoring pathways
• hypertension complications and multi-morbidity

Neurological

• Parkinson’s disease
• multiple sclerosis
• epilepsy
• dementia support pathways (capturing caregiver and functional changes carefully)

Autoimmune and inflammatory

• rheumatoid arthritis
• lupus
• inflammatory bowel disease
• vasculitis and other rare inflammatory conditions

Respiratory

• COPD
• severe asthma
• pulmonary fibrosis and other progressive lung conditions

Rare diseases

• any condition where small numbers prevent robust learning at a single site
• conditions where multi-country aggregation could unlock insights faster

The point isn’t to chase breadth. The point is to create a repeatable model: consistent data, respectful governance, and research acceleration.

How We Make Real-World Research Faster (Without Cutting Ethical Corners)

Speed is often misunderstood.

In health research, speed does not mean rushing consent, skipping validation, or taking shortcuts with privacy. It means removing avoidable friction and building systems that make the right work easier to do.

Here are the practical levers that can accelerate progress.

1) Standardised onboarding and follow-up

When every site collects different variables in different formats, research slows.

A shared platform can create:

• a core dataset that all partners align to
• condition-specific modules built with specialists
• follow-up schedules that are realistic and useful
• data dictionaries that reduce ambiguity

This improves signal quality and makes multi-site studies feasible without constant reinvention.

2) Capture what happens between appointments

A lot of deterioration, side effects, and supportive care needs occur between clinic visits.

Patient and carer input helps fill gaps:

• symptoms that fluctuate daily
• tolerability of treatment
• mobility and function changes
• mental health and fatigue patterns
• barriers to care (transport, cost, time)

This is not “soft” data. When captured consistently, it becomes vital context.

3) Reduce the research admin burden for clinicians and coordinators

If participation is administratively painful, it stays small.

Good design can:

• pre-fill fields where possible
• support structured templates for quick review
• integrate with existing workflows when feasible
• avoid duplicate entry
• provide immediate utility (summaries, longitudinal views, consistent reporting)

4) Build research access pathways that are repeatable

One of the slowest parts of multi-study work is repeatedly negotiating access, extraction, and approvals.

A platform can support:

• repeatable data use agreements
• auditable access controls
• role-based research workspaces
• consistent export formats for approved projects

This doesn’t reduce rigour; it reduces repetition.

5) Enable multi-site, multi-country collaboration responsibly

For rare cancers, rare MND subtypes, and complex multi-morbidity, collaboration is the difference between “not enough evidence” and “actionable signal”.

A JV-led infrastructure can support:

• federated collaboration models (where data stays within jurisdictions but insight is shared)
• common variable definitions
• shared protocols
• consistent reporting and quality standards

6) Turn supportive care into measurable knowledge

Foundations often do the hardest work: equipment programs, navigation, counselling, peer support, financial support.

But supportive care is frequently under-measured.

Capturing service use and outcomes can help:

• show what materially improves quality of life
• advocate for funding with evidence
• identify gaps and inequities
• refine programs with feedback loops

This is research acceleration in the truest sense: learning what helps people sooner.

The Principles That Make This Worth Doing

If a JV is going to be trusted by patients, clinicians, and the research community, it needs clear principles—written down, lived, and audited.

Here are the principles we design around.

Patient dignity and control

• informed consent that is understandable
• options to participate at different levels
• ability to withdraw
• transparency on data use

Privacy by design

• minimise data collection to what’s necessary
• strict access controls
• audit logs and monitoring
• secure handling and storage

Shared governance and accountability

• foundations and research centres at the table
• patient representation where appropriate
• ethics oversight and clear escalation pathways
• published policies for data access and usage

Scientific and clinical validity

• define hypotheses and endpoints clearly
• validate models appropriately
• avoid overclaiming
• publish methods and limitations transparently

Equity and inclusion

• design for diverse cohorts (not just well-resourced sites)
• reduce barriers to participation
• ensure insights don’t only represent privileged subsets

Practical value

If clinicians and patients don’t find it useful, it won’t last.

The platform must support:

• better conversations in clinic
• clearer care coordination
• easier participation in studies
• earlier supportive interventions

What “Everyone Benefits” Looks Like in Practice

A non-extractive, mission-led AI collaboration should create tangible benefit across the ecosystem. Not theoretical. Practical.

Patients and carers benefit through:

• clearer tracking of symptoms and function over time
• earlier identification of support needs
• improved access to trials and research opportunities (with consent)
• reduced need to repeat their story across providers
• stronger advocacy, because foundations can demonstrate unmet needs with evidence

Clinicians benefit through:

• a longitudinal view that makes complexity easier to manage
• consistent measurement tools and less fragmented information
• better ability to coordinate across specialties and services
• research participation that doesn’t add disproportionate admin load

Researchers benefit through:

• larger, more representative cohorts
• consistent variables and definitions
• longitudinal outcomes rather than snapshots
• faster feasibility checks and cohort identification
• reproducible data workflows and documentation

Foundations benefit through:

• stronger evidence for advocacy and policy change
• clearer demonstration of program impact
• improved ability to convene partners around a shared asset
• infrastructure that can support multiple studies over time
• an enduring platform that continues learning as the community grows

Health systems and communities benefit through:

• earlier intervention and fewer crisis events
• more effective supportive care pathways
• improved alignment between research and real-world practice
• better allocation of resources informed by actual needs

When done properly, the core outcome is simple: less time lost—for patients waiting, for clinicians searching, and for researchers rebuilding.

What We Bring at Digital One Agency (and Why a JV Structure Helps)

In this invitation, we’re not asking foundations and research centres to “buy software”. We’re proposing a partnership where each party contributes what they do best.

Digital One Agency contributes:

• secure platform engineering
• data architecture and pipelines
• integration capability (where permitted and feasible)
• applied AI development (NLP, longitudinal modelling, cohort tools)
• governance-by-design implementation (audit logs, access controls, consent workflows)
• usability design that reduces burden for patients and clinicians

A JV structure helps because it supports:

• shared governance
• long-term stewardship
• transparency around decisions
• sustainability beyond grant cycles
• a shared asset that can grow with the mission

The intent is to build capability with partners, not simply deliver a product.

How a Partnership Can Start: A Practical Path That Doesn’t Try to Do Everything at Once

A sensible start is focused and achievable.

Step 1: Define the mission-aligned use case

Examples:

• an MND longitudinal registry with consistent functional measures and PROs
• a cancer survivorship outcomes program with side-effect and late-effect tracking
• a rare cancer cohort discovery tool across partner centres
• a kidney failure registry linking dialysis parameters with wellbeing outcomes

Step 2: Co-design the minimum viable dataset (MVD)

Not “everything we can collect”—the smallest set that creates value.

• core demographics and consent
• clinical measures agreed by specialists
• a practical PRO set (short, meaningful, repeatable)
• key outcomes (hospitalisations, treatment changes, functional milestones)

Step 3: Governance and ethics scaffolding

• data access policies
• research approvals pathway
• participant communications plan
• audit and oversight mechanisms

Step 4: Build the platform module and pilot with a small cohort

• iterate based on real-world use
• measure data completion and usefulness
• reduce friction points
• establish quality baselines

Step 5: Scale partnerships and research use

• add sites or regions
• integrate additional data sources where feasible
• enable approved research projects
• publish early learnings transparently

This approach creates momentum without compromising trust.

A Note on Language and Expectations

AI does not “cure” cancer or MND. No responsible organisation should claim that.

What AI can do—when led ethically and supported by good data—is help shorten the distance between:

• observation and understanding
• understanding and action
• action and improved outcomes

It can help research and care communities learn faster, together, in ways that honour patients’ time and experiences.

That’s the ambition here: better learning systems, not miracle claims.

An Open Call to Foundations and Research Centres Worldwide

If you’re a foundation or research centre working in:

• cancers (common, rare, paediatric, haematological)
• motor neurone disease (MND)
• liver failure and severe liver disease
• kidney failure and chronic kidney disease
• and other serious, complex conditions where longitudinal learning can improve care and accelerate research

…we invite you to explore a joint venture partnership with Digital One Agency to build ethical AI research infrastructure that benefits everyone involved.

This isn’t a closed initiative. It’s intentionally open—because progress accelerates when communities collaborate, not when they compete in silos.

Contact Us — Let’s Talk

If this approach aligns with your mission, we’d welcome a conversation.

When you reach out, it helps to include:

• your condition focus (e.g., cancer subtype, MND program, transplant pathway, survivorship)
• your biggest research or coordination bottleneck right now
• the stakeholders you’d want involved (clinical lead, research lead, governance lead, patient/community lead)
• whether you’re looking for a new registry, an upgrade to an existing one, or a multi-site collaboration layer

Contact Digital One Agency and let’s talk about a JV partnership that builds the right foundations—so research can move faster, responsibly, and with real-world impact.