Water utility · twentieth century
For most of the twentieth century, water utilities operated as three islands. Reservoir A served District 1. Reservoir B served District 2. Reservoir C served District 3. Each had its own staff, its own schedule, its own pipes. They shared a name — not a system.
Water utility · when one fails
Now Reservoir B is contaminated. District 2 has no water — and the neighbouring reservoirs can't help because their plumbing doesn't reach. No shared treatment, no blending, no automation. Operators picked up the phone. That is the cost of silos.
Data world · Snowflake
Snowflake is the cloud data warehouse that unifies everything from Topic 2 onto one platform: storage, pipelines, catalogue, and lineage — all orchestrated as a single system. Every team draws from the same reservoir. No gaps between systems, no telephone calls to co-ordinate.
What this means
Snowflake brings the data onto one orchestrated platform: a single source of truth. Every team draws from the same data, each with its own dedicated computing power.
With no gaps between systems, problems can't hide in the cracks — and we can surface insights across different data sources.
Data world · monitoring
Monitoring is the attempt to be ahead of the wave. We want to know that things are not working before anybody else does. Using Snowflake allows us to detect issues even before we wake up.
What this means
Monitoring continuously watches all pipelines, transformations, and data quality. Did tonight's pipeline finish on time? Did a data source go silent? When something breaks at 2 a.m., an alert goes straight to the dedicated Teams channel.
Problems are caught before the first analyst opens an empty dashboard at 8 a.m.
Data world · quality control
Data quality control checks accuracy, completeness, consistency, and freshness of the data — not once at commissioning, but every single day.
What this means
Data quality control runs those same four checks automatically on every data load:
- Accurate — do the numbers match the source? (warehouse revenue = what was actually invoiced)
- Complete — is anything missing? (no blank fields, no missing days)
- Consistent — do systems agree? (CRM says Munich, billing says Frankfurt = a problem)
- Fresh — is it current? (a dashboard showing last week's figures as "today" is misleading)
It's the last line of defence between data and the choices people make with it.
Together
Snowflake orchestrates. Monitoring watches. Quality verifies. Together they turn a sprawl of disconnected systems into something a company can rely on — every minute of every day.