Bruin CLI

Appearance

Bruin - Firebase to GCP Template ​

This pipeline is a simple example of a Bruin pipeline for Firebase.

The pipeline includes several sample assets:

  • analytics_123456789/events.asset.yaml / events_intraday.asset.yaml: Sensors that watch for new Firebase export tables in BigQuery and trigger downstream tasks. Keep one depending on your export type (daily vs intraday).
  • analytics_123456789/parse_version.sql: BigQuery UDF that normalizes app version strings (e.g. 1.20.3 → 001.020.003) for sortable comparisons.
  • events/stg_events.sql: View over the raw analytics_*.events_* wildcard table. Owns all parsing/flattening logic (event_params, user_properties, experiments, device, geo).
  • events/events_json.sql: Materialized incremental table over stg_events, partitioned by dt, clustered by event_name + user_pseudo_id. Used for historical queries.
  • events/events.sql: View that unions events_json (history, dt <= end_date) with stg_events (intraday, dt > end_date) for near-real-time coverage without re-scanning history. Adds typed columns (screen, session, ads, idfa/idfv).
  • user_model/stg_users_daily.sql: Incremental table with daily user-level aggregates (sessions, ad/IAP revenue, first/last device & geo of day) from events.events.
  • user_model/users.sql: User-level table with install-time attributes and cohorted retention/revenue metrics (ret_d{1..90}, {metric}_d{N}).
  • user_model/users_daily.sql: Enriched daily rollup that joins stg_users_daily back with users to tag each daily row with install context, days_since_install, and nth_active_day.

For a more detailed description of each asset, refer to the description section within each sql asset. Each file provides specific details and instructions relevant to its functionality.

Setup ​

Add your connections and environments to the .bruin.yml file at your project root, not inside the pipeline folder. You can read more about connections here.

Here's a sample .bruin.yml configuration:

yaml
environments:
  default:
    connections:
      google_cloud_platform:
        - name: "gcp"
          service_account_file: "/path/to/my/key.json"
          project_id: "my-project-id"

Important Notes ​

1- Rename analytics_123456789 (folder + references in stg_events.sql) to your Firebase analytics ID. 2- Keep only events_intraday.asset.yaml or events.asset.yaml depending on your use case. We recommend events_intraday since streaming data is not bound by the 1M events/day limit. stg_events.sql defaults to the intraday sensor — update its depends: block and replace your-project-id in events.asset.yaml if you use daily export instead. 3- Review TODOs: events/stg_events.sql, events/events.sql, and user_model/stg_users_daily.sql contain TODO comments. These indicate sections where you should make adjustments based on your data and project requirements (analytics ID, user_id vs user_pseudo_id, app-specific event params and metrics).

Running the pipeline ​

Run these commands from the generated firebase pipeline directory. To run the whole pipeline:

shell
bruin run .

You can also run a single task:

shell
bruin run assets/events/events.sql

You can optionally pass a --downstream flag to run the task with all of its downstreams.

That's it, good luck!