## Metadata - Author: [[Benn stancil|Benn Stancil]] - Full Title:: The Rise of the Analytics Pretendgineer - Category:: #🗞️Articles - URL:: https://benn.substack.com/p/the-rise-of-the-analytics-pretendgineer - Finished date:: [[2024-06-01]] ## Highlights > dbt is functions without a framework. There are [official best practices](https://docs.getdbt.com/best-practices), but they’re more suggestions than guardrails—and nobody wants to read the manual anyway. As a dbt project grows, the only thing keeping us from turning it into a spidering web of entangled models and duplicative logic is our own discipline and design sense. Which takes skills many of us never learned, because we never *needed* to.[4](https://benn.substack.com/p/the-rise-of-the-analytics-pretendgineer/#footnote-4) ([View Highlight](https://read.readwise.io/read/01hz88g1n3gp08cecvaj8jyz47)) > That seems like Snowflake’s approach here: In exchange for money and the Snowflake sales team selling Hex and Sigma, the two companies will prioritize building a [Snowflake Native App](https://www.snowflake.com/en/data-cloud/workloads/applications/native-apps/) over a [Databricks Lakehouse App](https://www.databricks.com/blog/introducing-lakehouse-apps), or integrate more directly with [Snowpark Container Services](https://docs.snowflake.com/en/developer-guide/snowpark-container-services/overview), or *won’t* build anything that might dramatically lower how much a customer uses Snowflake. ([View Highlight](https://read.readwise.io/read/01hz88n7x0hj1nfk9qcjfs994p))