How orchestra-hq/orchestra-blueprints works
orchestra-blueprints is a collection of reference implementations for Orchestra-driven data platforms, not one integrated application. Its principal workflow is to define or trigger data pipelines, extract from external systems, load warehouse data, and optionally publish lineage or analyze platform usage. The README also documents warehouse-specific dbt projects, Azure ML assets, metadata ingestion, and reusable orchestration patterns. Some documented areas are not represented in the supplied file tree or sampled source, so their internal wiring is not inferred.
Main entry points: hubspot_pipeline.py HubSpot pipeline · linear_pipeline.py Linear pipelines · google_sheets_pipeline.py Google Sheets pipeline · google_sheet_ingest.py Native Sheets loader
Components 22
Orchestration patterns
- Multi-pipeline runner
patterns/run_multiple_pipelines/run_multiple_pipelines.py - Runner API client
patterns/run_multiple_pipelines/_pipeline_runner.py - Warehouse savings analysis
patterns/warehouse_savings/graph.py
Data ingestion
- dlt ingestion examples
python/dlt - HubSpot pipeline
python/dlt/hubspot_pipeline.py - Linear pipelines
python/dlt/linear_pipeline.py - Google Sheets pipeline
python/dlt/google_sheets_pipeline.py - Native Sheets loader
python/google_sheets_native/google_sheet_ingest.py
Warehouse projects
- Warehouse dbt projects
dbt_projects
Platform lineage
- Lineage extractors
python/lineage/sources/bigquery.py - Lineage publisher
python/lineage/publish_lineage.py - Landed metadata
python/lineage
ML and workers
- Azure ML pipeline
python/azure/ml/pipeline_script.py - Data preparation
python/azure/ml/scripts/prep_data.py - Model training
python/azure/ml/scripts/train.py - Metadata API example
metadata_api/run.py
Other components
- Platform operator
- Orchestra platform
- SaaS and Sheets
- dlt runtime
- BigQuery destination
- MotherDuck destination
Connections 33
- Platform operator → Multi-pipeline runner: runs
- Multi-pipeline runner → Runner API client: delegates
- Runner API client → Orchestra platform: starts runs
- Platform operator → HubSpot pipeline: runs
- HubSpot pipeline → SaaS and Sheets: extracts
- HubSpot pipeline → dlt runtime: loads via
- dlt runtime → BigQuery destination: writes data
- Platform operator → Linear pipelines: runs
- Linear pipelines → SaaS and Sheets: extracts
- Linear pipelines → dlt runtime: loads via
- dlt runtime → MotherDuck destination: writes data
- Platform operator → Google Sheets pipeline: runs
- Google Sheets pipeline → SaaS and Sheets: extracts
- Google Sheets pipeline → dlt runtime: loads via
- dlt runtime → BigQuery destination: writes data
- Platform operator → Native Sheets loader: runs
- Native Sheets loader → SaaS and Sheets: reads sheet
- Native Sheets loader → BigQuery destination: loads rows
- Platform operator → Warehouse dbt projects: uses
- Platform operator → Lineage extractors: runs
- Lineage extractors → BigQuery destination: queries metadata
- Lineage extractors → dlt runtime: loads extracts
- dlt runtime → Landed metadata: lands metadata
- Lineage publisher → Landed metadata: queries
- Lineage publisher → Orchestra platform: publishes assets
- Platform operator → Azure ML pipeline: submits
- Azure ML pipeline → Data preparation: dispatches
- Azure ML pipeline → Model training: dispatches
- Data preparation → Model training: passes prepared data
- Platform operator → Metadata API example: runs
- dlt ingestion examples → dlt runtime: uses
- Platform operator → Warehouse savings analysis: uses
- Warehouse savings analysis → Orchestra platform: analyzes API