Symphony

I was diagnosed with severe eczema two weeks after I was born.

My childhood was dominated by flare-ups every few days and immunosuppressive medicines that often had little or no effect. Regular appointments with dermatologists showed me what medicine could do, and inspired me to pursue it as a career.

I got into medical school and trained to become a doctor.

The reality was different from what I expected. A consultation captures the flare or problem in front of you. But it rarely shows the weeks at home before it including what changed, what a person tried and what made things worse. Time pressure makes that context difficult to collect, so the immediate problem is treated while the root cause often remains unclear.

Sadly, the trend in the NHS was that these time pressures were becoming ever increasing.

I joined the Academic Foundation Programme and continued the leadership work I had started at university. My project examined leadership and innovation inside the NHS, including work with local Academic Health Science Networks.

During the same period I joined the NHS Clinical Entrepreneur Programme and started building the foundations of EczemaDoc with my co-founders Dr Niall Jawad and Dr Noreen Akram.

The programme brought NHS staff together for mentoring, information days, business teaching, regulatory support and introductions to other founders. I worked alongside other NHS innovators while developing the early product and testing the problem it needed to solve.

My experience with eczema led me to co-found Eczemadoc, a tool for tracking symptoms, discovering possible triggers and managing flare-ups at home. It was released on the App Store and Google Play, then renamed Proton Health and expanded to acne and psoriasis.

The current product is Symphony. It combines photos, symptoms, lifestyle, mental wellbeing, diet and products in a 360° assessment, then looks for patterns and prepares a personalised plan. It supports people with eczema, acne, psoriasis, rosacea and other persistent skin concerns.

We received grants from Innovate UK in 2022 and the NIHR in 2023.

The grants funded separate pieces of research and development, including work on a smartwatch system for detecting itch and scratching. Each application required a defined clinical problem, development plan and measurable outcome before the work began.

KQ Labs is the Francis Crick Institute’s accelerator for data-led healthcare startups, and Proton Health’s data-driven approach made us a strong fit.

We joined the three-month cohort in winter 2022, which gave Proton Health tailored support including one-to-one mentoring, workshops, introductions and a final demo day. Alongside the KQ Labs funding, we raised further backing through StartUp Health and private angels to develop the product, test the commercial model and prepare the company for clinical partnerships.

Techstars is a global startup accelerator network; we joined its Physical Health Accelerator in Fort Worth, Texas, finishing with Demo Day in December 2024.

We interviewed 19 dermatologists and designed the workflow with them. Three clinics piloted the product. We pitched at Demo Day and Techstars invested in the company.

That work moved Proton towards a clinical co-pilot: collecting a patient’s history before the visit, summarising the important context and preparing the consultation for the clinician.

In winter 2024, we worked with UCB to co-create the hidradenitis suppurativa (HS) portion of their app.

HS is a painful, long-term inflammatory skin condition that is still poorly understood, often misdiagnosed and difficult to track and manage between appointments.

We were selected for the programme because our dataset and experience identifying triggers for skin conditions suited the problem well. We worked closely with the UCB team to co-create the HS portion of the app, culminating in a demo day at their headquarters.

Proton Health achieved ISO 27001 certification and Class I medical device registration.

We registered the Class I medical device with the MHRA. ISO 27001 followed a separate independent audit of our information security management system.

The work covered risk management, access controls, incident processes, supplier checks and the evidence needed to show that those controls were operating.

We white-labelled the app for selected partners and connected tracking and education with products people were already using at home.

We partnered with Gladskin, Codex Labs, Cosi Care and Eczema Clothing: listing their products in our marketplace and co-creating practical content for people managing eczema.

I developed a multivariate model to predict the next day’s skin score and identify increased flare risk.

The model used anonymised records from 130 people and combined symptom history, products, food, behaviour and environmental variables. It predicted the next day’s skin score, with a prediction counted as correct when it fell within a predefined tolerance range of the recorded score. The leading internal test produced an F1 score above 0.80.

The predicted score gave people an earlier signal to review likely triggers and act before symptoms became more severe.

Our proof of concept showed improvements in symptoms, mental health and people’s sense of control.

Our internal analysis of 40 members compared follow-up assessment scores with each member’s baseline. Average scores improved by 34% for symptoms, 48% for mental wellbeing and 58% for people’s sense of control over their triggers. These were product outcomes, not medically assessed trial results.

As of September 2026, the current Symphony product reports more than 11,700 assessments across more than 100 doctors.

A structured database for nutrition, lifestyle and guided mental-health interventions used in inflammatory skin research.

Problem

Symphony covered acne, eczema, psoriasis, rosacea and topical steroid withdrawal. Its research included food, lifestyle changes and guided mental-health sessions. The source material used different terms and levels of detail, so records could not be compared or queried reliably.

Solution

I designed a clinical ontology: a fixed set of fields and allowed terms that describe each intervention in the same way. Python pipelines extract records into that structure, validate them and write them to Airtable.

Steps

1. Define strict JSON schemas with 40 to 60 or more fields for each record type. Required fields and controlled choices prevent different words being used for the same concept.
2. Model nutrition records with compounds, inflammatory markers such as CRP, metabolic measures such as insulin response, gut-skin factors, allergens, histamine classification and dermatology outcomes.
3. Model behavioural interventions with dose, frequency, adherence threshold, washout period, contraindications, effort and rules for continuing or stopping a self-observation period.
4. Parse guided-session transcripts and label the methods used, including breathwork, cognitive reframing, somatic practice and compassion-based work.
5. Record proposed mechanisms, including vagal tone, HPA-axis activity and parasympathetic activation, alongside the symptoms or emotions targeted.
6. Split very large extraction jobs into separate passes for core details, targets and mechanisms. Merge the passes only after field-level validation.
7. Read the live Airtable schema through its metadata API. Map each value to the correct Airtable type, including selections, checkboxes, text and JSON fields.
8. Run records in batches with a separate error boundary for each item. One bad record is logged without stopping the rest of the batch.
9. Support safe reprocessing so a corrected source or schema can be run again without creating uncontrolled duplicates.

Results and benefits

Structuring a source manually takes about 25 to 35 minutes. Automated extraction followed by review takes five to eight minutes, reducing handling time by roughly 75%.

Nutrition, lifestyle and guided-session evidence can be searched using the same field rules. Validation catches missing fields before storage and failed records are isolated for correction.

I had spent years on strong immunosuppressive medicines. Tracking my symptoms in the Proton Health app found triggers that treatment alone had missed.

Two months of tracking linked strawberries, other berries and apples to my symptoms. I also switched my shampoo after finding an irritant ingredient, started using Adex Gel for its Niacinamide and trialled a handheld red-light device from Amare Skin. My last flare was in summer 2023.