< Government >
Leveraging Machine Learning for
Public Service Transformation
We extract insights from large government datasets to analyze service utilization, design optimal delivery models, and build real-time monitoring systems that support quality public service provision. Our system tracks users interactions, service delivery metrics, personnel performance, and resource allocation across hundreds of government facilities. We process millions of transactions yearly, with detailed service delivery information including demographic data and procedural outcomes.
Supporting Rwanda's Healthcare Reform with RSSB
Rwanda's Social Security Board (RSSB) plays a critical role in financing primary healthcare services. Causal Foundry is supporting the design, rollout and monitoring of a major health financing reform affecting the 1,000+ health centers and health posts in the country
Predicting Engagement & Demand
Casual Foundry has extracted insights from RSSB's large dataset of patient-level provider claims to analyze health service utilization and consultation costs, design the capitation formula, and build real-time monitoring systems to support quality healthcare provision.
Deep data analysis on the Intelligent Health Benefits System (IHBS)
- Tracks patient visits, diagnoses, services, personnel, and prescriptions across the 1,000+ PH C facilities since 2023
- 20m+ visits per year to 6m+ patients, with detailed patient-level service delivery information including diagnosis and 80m+ procedures, prescriptions, all tests per year
Data-driven design of the capitation model for Health Centers
- Developed a capitation formula that ensures a stable transition for health facilities, preventing system disruptions.
- The model has the flexibility to accommodate both basic and advanced health centers under a unified scheme
A monitoring system with outlier detection
- Tracked key indicators such facility utilization, service delivery, and test rates per visit.
- Implemented advanced algorithms to detect outliers and anomalies using time- series and cross-sectional analysis.