VytaloLet’s talk VYTALO / OUR APPROACH & SAFETY
Meet Vytalo and explore our approach to safety, evidence and responsible development.
OUR COMMITMENT TO YOU
Understand what informs your support, who reviews it and what still needs to be proven.
THE SYSTEM
Nutrition × intelligence × human coaching.
Food, health data and coaching should inform one another. Our vision brings them together around you.
Our nutrition approach brings condition-specific food choices and practical meal planning together with professional guidance. The care design considers your health needs, preferences and everyday routine, with registered-dietitian input where appropriate.
04 / THE CURRENT PHASE 2 APPROACH
Transparent rules help the care team decide what to review.
The initial design uses a rolling 14-day CGM window. A data-quality gate checks completeness; safety checks take priority over any weighted Green, Amber or Red review score.
Rolling CGM window
Data needed for scoring
Safety checks first
Insufficient data means no calculated score. Serious observed safety events still go for human review. The priority score is not a probability of disease.
Continue agreed support and check-ins.
A coach reviews the pattern and context.
Assess safety flags and escalation needs.
This is not an emergency monitoring service. Do not wait for a Vytalo review when urgent medical help is needed. Follow your existing care plan and your device’s safety guidance.
05 / A RESPONSIBLE ROADMAP
Predictive AI belongs on the roadmap only when the data, clinical evidence and governance support it. Development milestones are evidence gates, not promises of a launch date.
CGM-led review priorities, data-quality checks, safety overrides, structured coaching and outcome capture.
XGBoost is a candidate for future individual prediction, with SHAP to explain model outputs. LCA, LPA, LFA and association-rule mining remain exploratory research options.
Time-series forecasting, patient similarity, uncertainty estimation and advanced models may be considered as the evidence base grows.
Before any predictive model influences care: data-readiness assessment, validation against appropriate baselines, calibration and fairness checks, clinical sign-off, and shadow testing without affecting patient decisions. A 4–8 week shadow period is a planning assumption, not proof of readiness.
YOUR INFORMATION & YOUR CHOICES
These commitments guide development and must be tested before the service launches.
Understand what is collected and why. Manage sharing, request an export or deletion, and receive a clear explanation of any retention requirements.
Access limited to each person’s role, stronger sign-in for professionals, encryption and a record of who viewed or changed information.
Plain language, scalable text and assisted setup, tested with older adults and people with less confidence using technology.
Clinical goals need to suit the individual. Future AI also needs representative data, fairness checks and clinical validation before it can influence care.
CGM consensus, NICE guidance and clinical safety planning inform the approach. Research and requirements are documented; implementation and assurance work continue.
A rolling 14-day CGM review, data-quality gates, seven safety checks and structured coaching — with every clinical output reviewed by a person in the current design.
XGBoost and SHAP are candidates for individual prediction and explanation. Use would require representative UK data, validation, clinical sign-off and shadow testing.
Time-series forecasting, patient similarity and uncertainty estimation may be considered as the evidence base grows and their added value can be demonstrated.
06 / BUILT WITH PRIMARY CARE IN MIND
A staged NHS integration roadmap.
Vytalo’s ambition is to make glucose history, wearable context and coaching activity useful to clinicians, in a form that fits their work. Each stage depends on assurance, agreements and partner readiness.
ALL NHS INTEGRATIONS ARE PLANNEDA consented 14-day summary for the patient to share, or for an authorised care team to send through an approved secure route.
Gate: assurance review, clinical safety and data-sharing agreements.A planned read-only view of trends and coaching history, with professional identity checks and audited access.
Gate: pilot evidence, identity integration and information governance.A longer-term ambition for structured information to appear in GP systems, using appropriate FHIR standards and approved integration routes.
Gate: NHS partnerships, onboarding availability and accreditation.07 / EVIDENCE, SAFETY & GOVERNANCE
Published research informs the approach. Vytalo-specific performance must still be demonstrated. A literature reference is not evidence that the platform delivers a particular clinical outcome.
We are building toward NHS assurance requirements. This website makes no claim of completed DTAC assessment, NHS endorsement, medical-device certification or clinically proven predictive AI.
Discuss the evidence roadmap08 / ABOUT VYTALO
A connected approach to preventative health, built around the person.
Vytalo Limited is developing a platform that connects health data, practical lifestyle support and qualified human review. The current platform focus is prediabetes, Type 1 diabetes and Type 2 diabetes, with broader metabolic health considered in the longer-term vision.
Our ambition is to help people understand their everyday patterns and help care teams make better use of the information between appointments.
Ask about the planned experience, availability or working with Vytalo.