An end-to-end data pipeline for U.S. Bureau of Economic Analysis (BEA) data:

• Ingest economic releases from the BEA API
• Transform datasets with dbt
• Store in a SQL database (Postgres or DuckDB)
• Visualize in Apache Superset with prebuilt dashboards

This project is a hands-on demo of modern data stack tooling: ingestion + transformation + BI.

• 🔄 Automated data ingestion from BEA API (GDP, trade, PCE, etc.)
• 🧹 dbt models for cleaning, staging, and aggregating BEA releases
• 📦 Database integration with Postgres (default) or DuckDB
• 📊 Prebuilt Superset dashboards for macroeconomic analysis
• 🐳 Easy setup with docker-compose (Superset + Postgres + dbt)

superset-dbt-bea/
│── data\_ingestion/
│    └── bea\_downloader.py   # Fetch BEA data from API
│── dbt\_project/
│    ├── models/
│    │   ├── staging/        # Clean raw BEA data
│    │   ├── marts/          # Aggregated metrics
│    │   └── metrics/        # Growth rates, ratios, rolling averages
│── superset/
│    └── dashboards/         # JSON exports of example dashboards
│── docker-compose.yml       # For Superset + Postgres + dbt
│── README.md

• GDP growth over time
• Contributions by industry
• QoQ vs YoY toggle

• Imports vs exports
• CPI / PCE inflation trends
• Trade balance as % of GDP

👉 (Screenshots coming soon — export Superset dashboards as JSON + PNG and place them here!)

git clone https://github.com/YOUR_USERNAME/superset-dbt-bea.git
cd superset-dbt-bea

docker-compose up -d

This spins up:

• Postgres (data warehouse)
• Superset (BI tool)
• dbt (transformation engine)

python data_ingestion/bea_downloader.py

cd dbt_project
dbt run

# Log in to Superset at http://localhost:8088
# Import JSON files from superset/dashboards/

• Apache Superset – Visualization
• dbt – Data transformation
• Postgres / DuckDB – Storage
• BEA API – Data source
• Docker Compose – Local orchestration

• Add support for more BEA datasets (regional GDP, income)
• Publish dbt models as a package (dbt-bea)
• Automate Superset dashboard import
• Add CI/CD pipeline for refreshing data (GitHub Actions)
• Deploy demo online with lightweight database (DuckDB/SQLite + Superset in container)

Pull requests are welcome! Ideas for improvements:

• New dbt models (e.g., employment, housing data)
• Additional Superset visualizations
• Better dashboard theming

MIT License — feel free to use and adapt.