Resolyst: continuous specialty care

Consider a real patient journey. As a teenager she had mild scalp psoriasis, forgotten as a minor skin complaint. Years later she develops acute joint pain; her GP refers her to a rheumatologist. They run a battery of scans and blood tests, diagnoses psoriatic arthritis, starts an immunosuppressant, and books a review appointment 12 months later.

Months later, a sore heel develops. After a four week wait, she manages to get an appointment with the rheumatologist, who orders an MRI. The follow up rheumatology appointment confirms plantar fasciitis - but there is no clear pathway, so she googles a podiatrist and buys orthotics alone.

New bowel symptoms bring her to her GP, who refers her for a colonoscopy; her rheumatologist never hears about this. A joint flares, there is no GP slot, so she re-books another rheumatology appointment, and is handed NSAIDs.

One patient, four specialties, zero coordination. And underneath it all, a quieter gap: she is on a powerful immunosuppressant in a fast-moving drug class, but no expert panel is reviewing whether her management should change as the evidence evolves. Every decision sits with one local rheumatologist, however excellent, working alone. The atopic patient shows the same pattern even more starkly, with asthma, eczema and allergy, three biologically linked conditions, managed by three teams who never share a record.

This is not an accident. Healthcare is organised around the event, not the programme. You wait months to be seen, get a short appointment, and are discharged into silence until something goes wrong badly enough to re-enter the queue. But the disease never stops between appointments. Specialty care is defined by fragmented excellence. It spans more than 200 specialty silos in a $4tn outpatient market. The best specialists are oversubscribed and cannot scale, and there is zero continuity between visits.

Resolyst is built to fix the part everyone has accepted as broken: the continuous management of a condition between appointments and across specialties. We are not another symptom checker, single-condition app or digital front door. We are the clinical workflow layer that lets a specialist run a continuous, multidisciplinary care programme for a large book of patients, with expert input flowing in and clean escalation flowing back to the consultant and, when warranted, to an international panel of the leading experts in that cluster of conditions.

Multidisciplinary care should not be a privilege of the largest institutions

Genuine multidisciplinary care requires the coordinated input of specialist nurses, dieticians, pharmacists, physiotherapists and occupational therapists working around a consultant. Today, that care exists almost exclusively inside large healthcare organisations and academic medical centres. Only they have the scale to employ a full roster of allied health professionals and convene expert boards. The independent specialist and the small or medium clinic do not.

Resolyst elevates independent specialists to operate like a micro-academic medical centre, giving them virtual or in-person agentic and allied health professional care and a connection to international expert panels, without building any of it in-house. The specialist becomes a decentralised healthcare organisation of their own, and connected to the network of overlapping specialists and research organisations. The continuous, coordinated model once locked inside the largest institutions becomes available to any specialist anywhere.

Four layers, one continuous system

The platform is built as four layers:

• Layer 1 is the continuous care platform: AI-enabled intake, MDT monitoring between appointments, and escalation back to the specialist clinic when clinically warranted.
• Layer 2 is the international expert panels, modelled on tumour boards but built for chronic clusters of conditions.
• Layer 3 is the research engine: structured, longitudinal, real-world datasets and trial-ready registries generated as a by-product of care.
• Layer 4 is the pathway-specific clinical intelligence models trained on that data, which over time create the opportunity to bring specialty-grade care to primary care or direct to patients.

The rest of this post is mostly about Layer 1, because that is where specialty care actually breaks, and it earns the right to everything above it.

Clinical care, not practice administration

Most companies approaching specialty care optimise the operations of a practice: scheduling, CRM, medication reminders, billing. These sit alongside care rather than inside it; they make the event-based model run more smoothly but do not change the model itself - the stuff that happens to the patient between appointments or involved in the clinical decision making. Resolyst, in contrast, is built around the clinical care itself, evidence-based management of the condition over time and active collaboration with the latest research in that disease, and is judged on clinical quality and outcomes, not on how efficiently an invoice is raised.

This is why the multi-specialty point is not a feature but the whole premise. We are deliberately focused on the complex intersection of polysymptomatic conditions, where a single patient's disease spans several specialties and the hardest decisions sit between departments. The psoriatic and atopic patients above are not edge cases; they are the centre of gravity for chronic disease, and exactly where the event-based, single-specialty model fails completely.

Resolyst is building continuous care loops with two themes of components. Some are disease-agnostic, the scaffolding every pathway shares: structured intake and its branching engine, the escalation funnel from agent to allied health professional to consultant to expert board, the auditable decision trail, the monitoring cadence and routing. Others are disease-specific, the clinical content that has to be right for each condition: the pathway-specific questions, severity scores and PROs, red flags, validated content libraries and expert input. The agnostic scaffolding lets Resolyst enter a new specialty quickly; the disease-specific content makes the care world-class. Most platforms have one or the other. The combination is the point.

From episodic appointments to continuous programmes

The core move is simple to state and hard to deliver: replace the episodic appointment with a continuous, agent-supported programme for long-term conditions. Patients enrol after an in-person specialist diagnosis, so the programme never holds the differential diagnosis risk pre-diagnosis. From there the pathway flexes to the patient's needs rather than forcing them back into a queue, follow-up is decoupled from monitoring and flare management, and escalation happens when it is clinically warranted rather than on an arbitrary calendar or when a patient can access an appointment.

For the patient, this is a Formula 1 pit crew for their condition. The heel, the bowel symptoms and the joint flare are no longer disconnected events navigated alone, the asthma, eczema and allergy are no longer three teams without a shared record, and the immunosuppressant decision no longer sits with one rheumatologist in isolation: through the international expert panel, her management is reviewed against current best practice by the people defining it. For the clinician, care and revenue extend beyond the ceiling of clinic hours; the specialist focuses on the complex cases that need them in person while the programme handles the rest under their oversight.

The international expert panel, rebuilt for chronic disease

Closing that immunosuppressant gap draws on one of medicine's most consequential ideas: team-based decision-making. The multidisciplinary team meeting evolved from the tumour board, formalised by oncology and eventually made mandatory, and spread to heart failure, IBD and transplant. But the delivery architecture has barely changed in seventy years and has hit a capacity ceiling: MDTs cannot scale at volume, so systems now reserve them for the most complex cases and let everything else revert to protocol. Resolyst keeps the clinical principle - several experts, one case, a documented consensus - whilst removing the constraints that kept it inside hospital walls. In other words, a tech-enabled, agentic-chaired, virtual, continuous, asynchronous MDT meeting. The expert panels are:

• Cross-border, not local, curated globally from leading academic centres on reputation and clinical depth rather than availability.
• Condition-led, not specialty-led, built around the condition cluster (atopy, the HLA-B27 spectrum, systemic autoimmune disease) rather than a single department.
• Asynchronous, not scheduled: cases are submitted with structured intake, reviewed independently, and returned within days.
• Continuous, not episodic: panels sit above ongoing pathways, informing protocols and reviewing escalations with a structured, attributable second opinion.
• Local care delivery: the patient's care always stays with their own clinician; the panel surfaces input, it does not take over the relationship.

What comes back is a documented expert review, ready to action, with suggested investigations, treatment options and rationale. For the patient on an immunosuppressant, it is the difference between care that depends on whoever they happened to see first and care reviewed by the people defining best practice in their disease.

The clinical workflow, and where the agents fit

A specialist's capacity is constrained by time, most of which goes on routine work that does not need a consultant. Resolyst surrounds each specialist with a pathway-specific multidisciplinary team, accessed fractionally and virtually, drawn on as needed rather than employed. The workflow has a clear structure: the subspecialist allied health team manages the large majority of inbound need, the consultant focuses on complexity with roughly one in ten interactions escalating to them, and the international expert board handles the most complex, multi-organ cases, on the order of one in a hundred. Every tier escalates only if needed, so care is matched to acuity and the scarcest expertise is reserved for where it changes outcomes. This also gives clinicians what they actually want: an academic challenge, rather than shift-based volume work. Historically, seeing complex patients still required the clinician to do all the other work - this model could actually see a consultant doctor running fewer clinics, seeing less patients but the ones they see are more complex cases that really need their expertise.

There is an agentic layer to this, and it matters to be precise about it. We build specialty-specific agents, subspecialist allied health professional agents per discipline built from validated clinical content, plus orchestration agents that assemble data, chase missing information, chair a board discussion and return a structured decision. They handle the high-volume, structured work: intake, navigation, triage, routine queries and the logistics of diagnostics and monitoring.

Crucially, they operate with humans in the loop and on the loop. Early on, clinicians approve and edit every message; as the models mature, and regulatory approval is achieved, humans move from approving each action to overseeing the system and intervening on exceptions. Liability follows the decision, not the tool, and every clinical decision leaves a time-stamped, auditable trail. This is the difference between Resolyst and AI products that promise to replace clinicians: we are not automating the doctor away, we are letting a great specialist safely oversee a far larger continuous operation than they could alone. The EHR remains the source of truth, or system of record, which agents read from and write back to.

Clusters, and care that learns

Most digital health companies optimise a single lane, one condition, one protocol, which falls apart where complex patients live: at the intersections. So we organise around condition clusters that reflect underlying biology. Atopy, eczema, allergy and asthma across dermatology, immunology and respiratory medicine, is the wedge; the psoriatic cluster is a natural next step; from there the architecture extends to rheumatic and broader autoimmune clusters. The continuous-care workflow, the escalation funnel and the shared branching engine are built once and reused. This also maps onto how research is organised, since therapeutic areas in pharma already span multiple specialties, so connecting clusters into unified pathways creates research-ready patient populations.

Because every interaction in a Resolyst pathway is structured, the by-product of delivering care is something specialty medicine has always lacked: clean, longitudinal, research-grade data generated inside real care pathways. That turns each pathway into a learning system, the substrate for proprietary clinical intelligence models and for research partnerships whose revenue can subsidise the cost of care and massively expand access. That is the flywheel: better continuous care produces better data, better data opens research, research subsidises access, and broader access produces better outcomes and a stronger network. But the data and research are downstream. What earns the right to all of it is the clinical workflow at the core, the continuous care programme that helps an independent specialist scale their practice, brings real multidisciplinary input into ongoing management, and escalates cleanly to the consultant and the world's leading expert panels when a patient needs it. That workflow is what specialty care has been missing. It is what Resolyst is building.