Clinical decision support

Identify diabetes risk earlier. Follow every patient over time. Understand where intervention is needed.

Diabetes Foresight AI is a secure clinical decision-support platform that combines structured diabetes screening, laboratory-first classification, internationally recognised risk assessments, longitudinal patient follow-up and privacy-protected geographic intelligence.

Abstract illustration of a diabetes analytics dashboard with trend charts and an aggregated regional map

Secure patient identity management

Role-based clinical access

Row Level Security

Audited sensitive actions

Aggregated geographic analytics

No definitive diagnosis or treatment prescription

Platform overview

A single clinical workspace covering the patient record, the assessment, the follow-up plan and the operational picture.

Persistent patient record

One secure patient file with a stable patient code and multiple dated assessments.

Structured diabetes assessment

Capture demographics, risk factors, measurements and laboratory results through a guided workflow.

Laboratory-first clinical logic

HbA1c and fasting-glucose classification take priority over questionnaire risk scores when laboratory results are available.

Longitudinal follow-up

Track weight, BMI, waist circumference, HbA1c and fasting glucose across assessments and follow-up visits.

Geographic intelligence

Review aggregated diabetes burden, screening coverage and geographic completeness by residence and service location.

Supervisor oversight

Monitor clinical priorities, overdue actions, laboratory reviews, follow-up ownership and data quality.

Clinical intelligence grounded in structured evidence

What the platform applies today

  • FINDRISC risk questionnaire
  • ADA Diabetes Risk Test
  • Automatic BMI calculation
  • Laboratory classification from HbA1c and fasting glucose
  • Clinical-priority rules
  • Separate clinical-action, laboratory-retest and lifestyle-review dates
  • Longitudinal trend analysis across visits
  • Data-quality checks for missing clinical and geographic fields

Accurate result wording

  • Diabetes-range laboratory result — requires clinical confirmation.
  • Prediabetes-range laboratory result.
  • History of high blood glucose.

The platform currently provides rules-based clinical decision support and risk stratification. It does not claim that a machine-learning prediction model is active until that model has been developed and clinically validated.

Built for responsible predictive intelligence

The platform is being prepared for future validated capabilities. These capabilities are not active today.

Individual diabetes-risk forecasting

Geographic burden forecasting

Identification of follow-up priorities

Explainable risk factors

Confidence and uncertainty reporting

Operational recommendations for screening and prevention programmes

Planned and subject to data sufficiency and clinical validation

How it works

Click each step to walk through the clinician workflow.

Step 1 of 5

Search securely

Search for an existing patient using the authorised identity-matching workflow.

Geographic intelligence

Two separate privacy-protected map layers describe where patients live and where care is delivered.

Patient residence layer

Aggregated burden and coverage by region, city and district of residence.

Unique patients

Total assessments

History of high blood glucose

Diabetes-range laboratory results requiring confirmation

Prediabetes-range results

Normal laboratory results

High-risk patients

Unclassified assessments

Geographic completeness

Small-area results are suppressed when fewer than 10 unique patients are available. Patient homes, individual patient markers, identity numbers, names, employer names and postal codes are never displayed.

Built for each role

Healthcare professional

  • Register patients
  • Complete assessments
  • Review follow-up and trends

Supervisor

  • Monitor aggregated clinical and operational performance
  • Review overdue actions and geographic intelligence

Administrator

  • Approve professional accounts
  • Manage roles, capabilities and reference data

Security and privacy

Direct patient identifiers may be collected only through the authorised patient-identity workflow. Sensitive identity values are encrypted server-side, access is permission-gated and sensitive actions are audited.

  • Full identifiers are not exposed in dashboards, maps or operational CSV exports.
  • Geographic analytics use aggregated data only.
  • Patient home coordinates are not collected.
  • Users only access records permitted by their role and organisation.
  • The application is clinical decision support and not a substitute for professional clinical judgment.

Frequently asked questions

Turn diabetes screening into continuous, actionable intelligence.

Tell us how you will use the workspace and we will point you to the right starting point.

Sign in to the workspace