Hello , we are planning to use DBT to transform raw data (JSON) in to dimensional model (1 Fact & 6-7 dimension) on Apache Pinot for our analytical use case. I see that there is no adaptor available in DBT for Apache Pinot. I would love to hear from community if anyone have integrated DBT with Apache Pinot by building customer adaptor or any other ways.

The sql executed on Pinot is meant for user facing applications. Dbt sql is meant for data transformations in a data warehouse. If you have a data warehouse, use dbt against that then import the results to Pinot for fast low latency queries that data warehouses aren’t optimized for.

Thanks for your response. Actually we don't have separate data warehouse. Data into Pinot (dimensional model) would be flowing through open table format (Apache Iceberg). In Iceberg, data will be raw data format (Parquet) & we are considering DBT to create tranformation model to convert into dimensional data model & execute on Pinot. On other side the we are planning to have universal semantic layer( like Headless BI) on top of Pinot. Any insights on approach to execute DBT Tranformation Models on Pinot will be of great help!

Execute DBT on iceberg then import into Pinot

Personally I would use DuckDB: https://duckdb.org/docs/extensions/iceberg.html Or MotherDuck, if you don't want to operate it.

Duck DB looks interesting, Jan. Appreciate your response. Will evaluate further to check feasibility with large data sets.