Working with Make Targets¶

This guide explains the Makefile structure and how to use the available targets across components.

The Makefile pattern¶

Each component has its own Makefile with operational targets. There is no root Makefile. You always run commands from within a component directory or use make -C <component> from the repository root.

Running targets¶

# From the repository root
make -C hub deploy-infra
make -C mock_data seed-shared-data
make -C projects/pudo deploy-schema

# Or from within the component directory
cd projects/pudo
make deploy-schema

Hub targets¶

Target	Command	Description
`deploy-infra`	`uv run python scripts/deploy_infra.py`	Bootstrap all Snowflake platform infrastructure.

Mock data targets¶

Target	Command	Description
`seed-shared-data`	`uv run python scripts/seed_shared_data.py`	Initial bulk load of PUDO data.
`add-morning-data`	`uv run pudo-generate add-day morning`	Simulate morning arrivals.
`add-evening-data`	`uv run pudo-generate add-day evening`	Simulate evening completions.
`simulation-status`	`uv run pudo-generate simulation-status`	Show simulation state.
`reset-simulation`	`uv run pudo-generate reset-simulation`	Reset simulation clock.

PUDO project targets¶

Deployment targets¶

Target	Command	Description
`deploy-schema`	`uv run python scripts/deploy_schema.py`	Create project schema.
`deploy-feature-store`	`uv run python scripts/deploy_feature_store.py`	Register entities and feature views.
`deploy-training-dag`	`uv run python scripts/deploy_training_dag.py`	Deploy training task graph.
`deploy-inference-dag`	`uv run python scripts/deploy_inference_dag.py`	Deploy inference task graph.

Execution targets¶

Target	Command	Description
`run-training-dag`	`uv run python scripts/run_training_dag.py`	Execute training pipeline.
`run-inference-dag`	`uv run python scripts/run_inference_dag.py`	Execute inference pipeline.

CLI targets¶

Target	Command	Description
`run-inference`	`uv run pudo-inference run`	Run batch inference via CLI.
`evaluate-predictions`	`uv run pudo-inference evaluate`	Evaluate predictions vs actuals.
`inference-alerts`	`uv run pudo-inference alerts`	Check alert conditions.
`inference-summary`	`uv run pudo-inference summary`	Print prediction summary.

Common patterns¶

First-time setup¶

make -C hub deploy-infra
make -C mock_data seed-shared-data
make -C projects/pudo deploy-schema
make -C projects/pudo deploy-feature-store

Training cycle¶

make -C projects/pudo deploy-training-dag
make -C projects/pudo run-training-dag

Inference cycle¶

make -C projects/pudo deploy-inference-dag
make -C projects/pudo run-inference-dag

Daily simulation cycle¶

make -C mock_data add-morning-data
make -C projects/pudo run-inference
make -C mock_data add-evening-data
make -C projects/pudo evaluate-predictions
make -C projects/pudo inference-alerts
make -C projects/pudo inference-summary

Adding new targets¶

To add a new target to a component's Makefile:

.PHONY: new-target

new-target:
    uv run python scripts/new_script.py

Follow the existing pattern:

Use .PHONY for all targets.
Use uv run to execute scripts (handles the virtual environment).
Keep targets atomic. Each target does one thing.