Ad-Tech Demos

Ad-tech only. Databricks Asset Bundle for demos (pipelines, data gen, ML). Use for workshops or as reference. More demos will be added over time.

Current demo: impressions — attribution (campaigns → impressions → clicks → conversions), Auto Loader, benchmark data, ML pipeline.

---

What's here

• Bundle: databricks.yml at repo root; each demo’s pipelines and jobs live in that demo’s resources/ (e.g. impressions/resources/*.yml).
• Impressions demo: impressions_etl (batch pipeline), autoloader_demo (stream JSON → Bronze), data gen notebook, ML pipeline (01–03), DS Agent notebook. Code under impressions/.
• Future: Additional ad-tech demos will live in their own folders alongside impressions/.

---

Quick start

Prerequisites: Databricks CLI configured (or set workspace.host in databricks.yml), and a Unity Catalog schema (e.g. agent_spark.genify_pipeline_dev).

# From repo root
databricks bundle validate
databricks bundle deploy

# Run pipeline
databricks bundle run impressions_etl

# Or run as job
databricks bundle run impressions_etl_pipeline
databricks bundle run generate_impressions_benchmark_data
databricks bundle run ml_pipeline
databricks bundle run ds_agent_notebook_00

---

Repo layout

ad-tech-demos/
├── databricks.yml     # bundle root; includes impressions/resources/*.yml
├── impressions/       # first demo: attribution, pipelines, data gen, ML
│   ├── resources/     # pipeline + job definitions (impressions use case)
│   ├── docs/         # demo guide
│   ├── notebooks/    # generate_impressions_data.py, 01–03 ML, DS Agent
│   ├── pipelines/    # impressions_pipeline (batch), autoloader_demo (streaming)
│   └── schemas/      # 00_baseline, 01_tier1_comments (Genie)
# More ad-tech demos (e.g. bidding, reporting) will be added as sibling folders.

---

Bundle targets

• dev (default): agent_spark.genify_pipeline_dev
• prod: agent_spark.genify_pipeline

Set workspace.host or workspace.profile in databricks.yml for your workspace.

---

Docs (impressions demo)

Topic	Where
Demo guide	impressions/docs/README.md
Batch pipeline	impressions/pipelines/impressions_pipeline/README.md
Auto Loader	impressions/pipelines/autoloader_demo/README.md
Data gen notebooks	impressions/notebooks/README.md
Schemas	impressions/schemas/00_baseline/, 01_tier1_comments/