23% Quality of Life Improvement: AI/ML Health Intelligence Platform for NHS Cancer Care
Key Metrics
23%
QoL improvement & 22% higher activity via AI-personalised care.
60%
Lower anxiety & 40% less distress over 10-week AI coaching.
INDUSTRY
HealthTech & Digital Health
SERVICES
AI/ML
— Custom ML Solution
Data & Analytics
— Unified Data Platform
DELIVERABLES
Computer Vision Expression Analysis Model
Personalised Risk-Stratification Engine
Wearable & Patient Data Pipeline
MLOps Training & Deployment Pipeline
Clinical Intelligence Dashboard
Overview
Built Alvie's computer vision expression analysis system and personalised risk-stratification engine, synthesising wearable data, symptom logs, and video consultation signals into dynamic health programmes. The platform delivered a 23% improvement in patient quality of life and 60% reduction in anxiety scores.
The Challenges
Cancer patients generate continuous streams of health data across wearable devices, symptom logs, and video consultations — but no intelligence layer existed to synthesise these signals into actionable clinical risk scores, leaving health coaches making decisions on fragmented, manually reviewed information.
Delivering personalised, risk-stratified health programmes at scale required continuously learning from individual patient data across physical activity, sleep patterns, mood, and clinical outcomes — capabilities far beyond what manual coaching or rule-based systems could provide.
Solution
Built a computer vision-powered expression analysis system using custom ML models to identify emotional and physical health indicators during video consultations — enabling proactive risk identification and clinical intervention before conditions escalate between appointments.
Engineered a data science pipeline ingesting continuous wearable data from Apple Health, Google Fit, and Fitbit alongside patient-reported outcomes — synthesising sleep patterns, heart rate, activity levels, and psychological scores into personalised, risk-stratified health programmes updated dynamically per patient.
Developed an MLOps pipeline for continuous model training, validation, and deployment — ensuring AI models adapt to evolving patient cohort data whilst maintaining clinical accuracy and NHS-grade explainability and governance requirements.
Built a clinical intelligence dashboard surfacing AI-generated patient risk scores, progress trends, and intervention recommendations across the full coaching cohort — giving health coaches data-driven prioritisation across every patient in real time.
Value Delivered
#01
Clinicians gain AI-generated risk signals from video consultations, wearable streams, and patient-reported outcomes — intervention decisions grounded in continuous data rather than periodic appointments.
#02
Each patient receives a dynamically updated health programme adapting to their actual physical activity, sleep, and psychological data — not a static plan issued at onboarding.
#03
Health coaches manage larger patient cohorts without losing personalisation — AI surfaces patients who need attention before they self-report a problem.
#04
Full ML governance and clinical audit trail embedded into every model decision — meeting NHS explainability and regulatory requirements without manual documentation overhead.
Key Metrics
23%
Improvement in patient quality of life (EQ5D-5L) alongside 22% higher physical activity scores — driven by AI-personalised programmes adapting to continuous patient data.
60%
Reduction in anxiety scores and 40% reduction in distress levels — validated across the full 10-week AI-guided coaching journey.
80 NPS
Exceptional patient satisfaction demonstrating clinical-grade AI personalisation delivered at scale.
Client Testimonial
"I have worked with Abdeali since starting in my current role and would highly recommend him as a partner. We are a fast moving start-up and Abdeali is playing a crucial role in ensuring our product is able to meet the changing needs of our business. I find him to be an excellent partner, providing a great connection between us and his offshore team with excellent communication skills and dedication. On a personal level it is a pleasure dealing with Abdeali – he has a good temperament with a can-do attitude. He is dedicated and diligent, going the extra mile to take calls, oversee releases or respond to queries even at unsociable hours, and leads his team by example to deliver when he says he will which is really important for me as we have users and customers that depend on this!"
Martyn
Head of product and compliance
