Self-Serve Metrics Stack (dbt-style with DuckDB + Prefect + Superset-ready)

A lightweight Analytics Engineer / Associate Data Engineer demo: events → models → scheduled builds → charts.

Contents

• seeds/events.csv — synthetic product events (visit, signup, activate, purchase)
• models/ — DuckDB-compatible SQL (staging + marts: sessions, funnel, retention)
• flows/flow.py — optional Prefect flow to run models
• scripts/run_once.py — builds the models and exports outputs/*.csv and charts
• images/ — generated charts (after running the script)

Quickstart (local)

python -m venv .venv && source .venv/bin/activate
pip install -r requirements.txt
python scripts/run_once.py

Screenshots