# Demo - Exploring the TPC-DS Dataset
This demo uses the popular TPD-DS dataset for transactional database benchmarking. You can read more about Trilogy and this benchmark tool here hereopen in new window .
No benchmarking right now, though - we just care that DuckDB can natively produce a nice representative data warehouse for TPC-DS with a built in extension, and that this dataset gives us a good way to explore the language syntax.
This demo uses an in-memory DuckDB database that may take some time to start for your first query. Trilogy is database agnostic; the same syntax will work on any backend such as Postgres, Bigquery, or Snowflake.
That's a lot to work with! Fortunately, someone has already defined a Trilogy model for TPC-DS hereopen in new window ; we'll be using that directly. We'll discuss models in more detail in a little bit.
To start out, let's focus on these Trilogy imports. You'll notice that these map closely to fact tables; that's not a coincidence. It's typical for a user query to be centered on one or more facts, so those are natural places to start a modeling hierarchy. Trilogy imports let you expose many entrypoints into the same relationship of tables to support intuitive query patterns.
