I pushed plotly dash to its breaking point to overhaul an absurdly slow data analysis system.

TL;DR using plotly for entire database analytics and making PDF reports to replace excel

The context here is long, horrifying, and kind of funny, but this might interest some. So my company has a few hundred clients that load data a few times a year and pay for analytics reports. Ok fine, just use airflow to ETL the transactions onto an analytical db, do whatever regressions/averages/percentiles you need to, use Looker to pull tables and figures into csv, xlsx, pdfs, …

analysis breaking dash data data analysis datascience plotly