Platform Engine

Data Diversity & Fairness Auditor (DDFA)

Independent algorithmic bias detection and data diversity governance for clinical AI systems.

Schedule a DDFA Demonstration Assess AI Maturity (EAMA)
Core Engine

Intersectional Bias Auditing on De-identified Historical Data

Healthcare AI models frequently inherit historic disparities embedded in clinical training data. DDFA is Synod IntelliCare's core auditing engine, engineered to identify intersectional bias across complex demographic variables including age, gender, race, geography, and socioeconomic determinants.

Operating via our Sidecar Protocol, DDFA conducts deep retrospective analysis on historical EHR data extracts without requiring live EHR integration queues or IT disruption.

Key Specifications

  • ✔ Intersectional Disparity Mapping: Evaluates combined demographic subgroups rather than single-variable metrics.
  • ✔ Sidecar Protocol: Operates on de-identified historical data extracts to decouple governance readiness from IT backlogs.
  • ✔ Regulatory Alignment: Produces documented subgroup performance evidence relevant to ACA Section 1557, Colorado AI Act and EU AI Act. Not a compliance determination.
  • ✔ Retrospective Auditing: Audits model outputs on de-identified CSV and JSON extracts, team-operated, with no EHR integration required. HL7 FHIR R4 middleware is on the 2027 roadmap.

DDFA Technical Architecture & Data Pipeline

DDFA Simplified Architecture Diagram

Figure: Sidecar retrospective audit architecture (available today) and planned HL7 FHIR R4 middleware (2027 roadmap).