Custom development · open source data stack

The analysis happens where the work happens.

No export to Excel, no separate BI project. Figures appear in the same application where the work happens — backed by PostgreSQL, Trino and Superset.

WHY DATA BELONGS IN THE APPLICATION

If analyses live in another tool, nobody looks. The insight arrives a day late and in a format that must first be interpreted. Data works where the decision is made: in the screen someone is working in.

How we proceed

1

Collect the questions

Which decision should this figure support. Without a question you get a dashboard that looks good and helps nobody.

2

Build the model

Clean tables with dbt, versioned and tested. A metric is defined once and identical everywhere — no more three truths.

3

Speed up queries

PostgreSQL for everyday use, TimescaleDB for time series, DuckDB or Trino for large analyses. Every load gets the right tool.

4

Make it visible

Embedded charts in the application, Superset for free-form analysis. Permissions apply as everywhere — nobody sees more than allowed.

WHAT THIS ACTUALLY DELIVERS
One

definition per metric — no more debate about which figure is right.

In flow

instead of in an export: the figure sits where the work is, not in a file.

Open

stack: PostgreSQL, dbt and Superset run anywhere — on-prem as in the cloud.

Technologies

All open source — no licence costs, no lock-in.

PostgreSQLTimescaleDBDuckDBTrinodbtApache SupersetApache KafkaGrafana

Related

UW
YOUR DIRECT CONTACT
Uwe Winnwa
Sales · cloud37 AG

30-min intro call — free.
No standard pitch. I listen, ask the right questions and give an honest assessment — even if we are not the right partner.