Platforms are building knowledge layers. Why you should own yours
Databricks and Snowflake now put a model of the business inside their platforms. For organisations on several platforms, a model built on open standards serves every one of them.
Outline, not yet written. This seed article shows how Insights render. Replace it in Sveltia CMS before launch.
The data platforms are building their own knowledge layers. Databricks’ Genie Ontology has been in preview since June 2026, and Snowflake announced Horizon Context the same month (Databricks, Typedef). Each puts a model of the business inside the platform, where it serves that platform’s AI.
When a platform’s layer suits you
- You run on one platform, and expect to stay there.
- The questions you ask stay inside that platform’s data.
When to own it on open standards
- You run on more than one platform, or expect to.
- You work in a regulated sector or in government.
- You want your model of the business to outlive any one vendor.
The industry is moving the same way
The specification for sharing measures between tools, launched as the Open Semantic Interchange, is now incubating at Apache as Ossie (Apache Ossie).
What we do
We build the same kind of layer on W3C standards, on your choice of platform, and you own it. Changing platform always costs something; our job is to keep that cost low.