growthepie improves transparency across the Ethereum ecosystem by curating onchain, offchain, and community-sourced signals into actionable metrics, dashboards, and research.
The gtp-backend repository powers growthepie’s public dashboard and data products. It runs on Python 3.10.12 and houses:
- Data ingestion adapters that pull metrics from RPC nodes, Dune, CoinGecko, L2Beat, DeFiLlama, BigQuery, and bespoke indexers.
- Curation and enrichment jobs that normalize raw payloads, stitch cross-source identifiers, and enforce common schemas before persisting to PostgreSQL.
- Analytics layers that transform curated tables into pre-aggregated highlights, trendlines, and API payloads used by the dashboard, alerts, and JSON exports.
- Airflow orchestration for scheduling hourly/daily DAGs, data quality checks, and downstream notifications.
backend/src/adapters/: Source-specific connectors (adapter_stables.py,adapter_defillama.py, etc.) plus shared clients likebigquery.py.backend/src/queries/: Jinja-templated SQL powering metrics (e.g.,select_txcount.sql.j2) and API endpoints.backend/airflow/dags/: Production DAGs grouped by feature area (metrics_*,api_*,other_*) with alerting and JSON generation pipelines.backend/src/api/: JSON builders (json_gen.py,json_creation.py) and the lightweight API surface inoli/api.backend/src/config.pyand friends manage environment-specific wiring.
- Ingest – Adapters fetch fresh slices of chain activity, market data, or third-party metrics.
- Transform – Helper modules clean, reconcile, and upsert into curated PostgreSQL tables.
- Analyze – SQL templates and Python processors compute KPIs for quantities like TPS, fees, security, and ecosystem health.
- Distribute – Airflow DAGs trigger JSON generation, API refreshes, alerts, and highlights consumed by growthepie.com and partner feeds.
Explore the rest of the growthepie org for front-end, infra, and research companions.