“Get data. Analyze it”. For this a system needs to query the data, and for this it needs a query language. Analysts, dashboards and policies require a syntax for queries for certain data sets. SQL is often the default tool for this, albeit having flaws and problems. At Armis we decided to build our own query language.

This decision has pros such as abstraction for our users, descriptive consistency across the system and avoiding lock-in to database systems. It also has cons such as development overhead, deep proprietary bugs and constant integration pains.

In my talk I’ll go through the decision to build our own query language, the hardships along the way, the fruits that it bore us.