I love how this goes to the logical end game of Product Analytics, instead of trying to do it in a dedicated product just reuse the modern data stack (PostgreSQ

I love how this goes to the logical end game of Product Analytics, instead of trying to do it in a dedicated product just reuse the modern data stack (PostgreSQL, Snowplow, DBT, Pandas). This way you have all the freedom and you don't have to wait for new data when you have a new idea to test since you can use historical data. Suggestion: instead of saying APLv2 just say Apache 2, that prevents confusion with the very different AGPL license.