BPJS Kediri Data Connector

This repository is the operational data connector for the JKN analytics research environment.

It is responsible for:

• Authenticating to Google Cloud (BigQuery)
• Executing the data warehouse pipeline
• Building analytic datasets used by research projects
• Exporting audit datasets for field auditors

This repository does not contain research analysis. All statistical models and research outputs live in separate repositories.

---

How the system works

This project runs the data platform pipeline:

Local R environment → BigQuery warehouse → Analytic marts → CSV audit export (Cloud Storage)

The warehouse SQL transformation logic is maintained in:

bpjs-chronic-analytics repository.

This repository only executes the pipeline.

---

Running the pipeline

1. Open the R project (bpjs_kediri.Rproj)
2. Authenticate to Google Cloud
3. Run:

source("pipeline/run_data_platform.R")

The pipeline will:

• Build raw tables
• Build staging tables
• Build analytic marts
• Generate audit indicators
• Export audit CSV files to Cloud Storage

---

Security

No patient-level data is stored in this repository.

Authentication credentials are excluded using .gitignore.

---

Related repositories

• bpjs-chronic-analytics – data warehouse SQL transformations
• jkn-obstetric-fraud-detection – obstetric claims research
• jkn-strategic-purchasing-ews – early warning system modelling

---

Research Ecosystem

This repository is part of a larger JKN research framework:

• jkn-claims-warehouse: transforms raw claims into analytic datasets
• jkn-obstetric-fraud-detection: applies the framework to detect abnormal delivery patterns
• jkn-strategic-purchasing-ews: early warning indicators for purchasing and policy monitoring

The repositories are intentionally separated to distinguish infrastructure, methodology, and research applications.