This is awesome
@Sidhant Gupta, I like the outline approach you have.
I wrote more from a BI perspective and that the semantic layer is nothing new here since 1991 since SAP BO and each BI tool has it. Which also makes it easier to understand. The key part for me, if you have multiple BI tools, data apps or notebooks consuming the same KPIs, is to centralize them in a semantic layer as YAML, ready to be checked into git or version control. Or the other use case when you need advanced features such as an API (graphql?), data modeling, or access permission, instead of starting from scratch, semantic layer solutions help you with this.
I wrote a
deep dive on our blog or you also find a lot of related relevant content on our glossary about the
semantic layer (check backlinks and graph).