Forma Health is Swept Certified
Forma Health uses Swept to validate the AI in its rare disease clinical trial platform. Swept probes the system with synthetic data to find hidden bias and privacy issues before they reach research data, and supervises it continuously in production, watching single interactions and wider patterns for drift. This dashboard shares what Swept checks.
- Adaptive testing
- Continuous supervision
- Bias and privacy checks
- Transparent reporting
Summary
- Checked continuously
Adaptive testing
Swept probes the system with synthetic data built for clinical trials, looking for behavior a fixed test suite would miss.
- Checked continuously
Bias testing
Answers are checked for bias across patient and caregiver groups, such as in how pain scales are read.
- Checked continuously
Patient privacy
How the system handles patient identifiers is tested, so identifying details stay protected.
- Checked continuously
Interaction monitoring
Single interactions and patterns across many patients are both watched in production.
- Checked continuously
Drift detection
Small shifts, such as in how the system reads medical terms, are caught before they skew research results.