Skip to content
CloudMagicTalk to us

Logistics

Illustrative example

A governed logistics lakehouse on Databricks and Google Cloud

A logistics operator

One governed catalog over freight, telematics, and warehouse data

12+
source systems unified under one governance model
Industry
Logistics / freight
Scale
12+ source systems
Platform
Databricks on Google Cloud
Engagement
Governed lakehouse

The challenge

  • Freight, telematics, warehouse, and EDI data each lived in its own silo
  • Analysts copied data between systems, and nobody trusted which copy was current
  • There was no single way to control who could see what
  • ETA predictions ran on stale, partial data

What it had to do

  • Bring every data source into one governed platform
  • Enforce access and lineage from a single place
  • Let BigQuery users keep working without moving off their tools
  • Support both SQL analytics and ML on the same tables

What we built

CloudMagic built a lakehouse on Databricks on Google Cloud, with Delta Lake tables on Cloud Storage and Unity Catalog governing access and lineage across every source. Analysts query the same governed tables from BigQuery through lakehouse federation, so they keep their existing tools. Looker sits on top for shared dashboards, and the ETA models run on the same data, no copies.

One lakehouse

Delta Lake tables hold freight, telematics, and warehouse data together.

Central governance

Unity Catalog controls access and tracks lineage across sources.

BigQuery federation

Analysts query governed tables from BigQuery, no data movement.

SQL and ML together

Dashboards and ETA models run on the same tables.

Reference architecture

Sources

  • Transport management
  • Telematics feeds
  • Warehouse & EDI

Ingest

  • Databricks Workflows
  • Pub/Sub

Governed lakehouse

  • Delta Lake
  • Cloud Storage
  • Unity Catalog

Consume

  • BigQuery federation
  • Looker
  • ETA models

Results

  • 12+ source systems now sit under one governance model
  • Analysts query governed data from BigQuery without copying it
  • ETA predictions run on current data and got noticeably more accurate
  • Access control and lineage live in one catalog, which shortened audits
  • A new governed dataset ships in days, not weeks, because the data is already there

For analysts

One trusted set of tables, reachable from the tools they already use.

For operations

ETAs built on current data, not last night's export.

For governance

Access and lineage controlled and audited from one catalog.

  • Databricks on Google Cloud
  • Delta Lake
  • Unity Catalog
  • Databricks SQL
  • Databricks Workflows
  • MLflow
  • Cloud Storage
  • BigQuery
  • Looker
  • Pub/Sub

Facing something similar?

Tell us where you are now. A senior engineer replies, usually within one business day.

Talk to a customer engineer