Analyses Not Queries
Self-service data analysis is not about running queries.
Queries are for fetching data. Fetching data is not analysis. It is retrieval — useful but different. Queries are a poor fit for analysis problems even when they (in theory) could be used to do an analysis. Ask any data scientist who has hacked together linear regression in SQL for some dashboard (a rite of passage).
Despite this fact, many self-service data analyses use tools like Tableau and Looker which, aside from their visualizations, are essentially query machines. The flexibility which data scientists could incorporate into self-service analysis is hampered by the limitations of these platforms. Unfortunately, there is usually nothing else that the data scientists can turn to without writing their own frontend.
Memo_er lets data scientists build self-service analyses using Python, a full programming language, which enables them to automate even complex analyses.