docs / workflows

Gold → BI / ML

Business intelligence dashboards and machine learning models consuming the Gold star schema. Role-based access control per apartment.

01Gold star schemasource

BI tools connect directly to the PostgreSQL Gold schema via the native PostgreSQL connector. The star schema is optimized for OLAP queries — dimension joins, time-series aggregation, filtering by apartment/room/device.

ConnectionPostgreSQL connector → gold schema → direct query on fact + dim tables

dashboards
02BI Dashboardsplanned

dashboard

tool

fact table

key metrics

Energy

Power BI

fact_energy_minute

Power (W), energy (kWh), cost (CHF), per device per room

Environment

Power BI

fact_environment_minute

Temperature, humidity, CO2, noise, pressure, door/window status

Presence

SAP Analytics Cloud

fact_presence_minute

Motion count, presence flag, door activity per room

Sensor Health

Power BI

fact_device_health_day

Battery levels, uptime, error counts per device

security
03Row-level security (RLS)planned

Each apartment owner only sees their own data. Implemented via Power BI RLS roles filtering on apartment_key.

role

filter

access

role_jimmy

apartment_key = (jimmy)

Own apartment only — energy, environment, presence

role_jeremie

apartment_key = (jeremie)

Own apartment only — energy, environment, presence

role_admin

none

All apartments, all dashboards, sensor health

ML
04Machine Learning modelsin progress

Energy consumption prediction

#27in progress

Forecast energy usage per apartment, considering weather data and historical patterns. Built with KNIME workflows.

Room presence prediction

#26in progress

Predict room occupancy from motion/door sensor patterns, output to presence_prob. Built with KNIME workflows.

StackKNIME Analytics Platform. Models will output predictions into gold.fact_ml_predictions (future table).