How comfortable are you with SQL, and what do you use it for as an analyst?

Comfortable enough to answer my own questions: joins, aggregates, subqueries, and window functions when I need ranking or running totals. I use it to size problems, verify what stakeholders tell me, find data quality issues before they wreck a migration, and build the baseline numbers behind a business case. Not tuning production queries, but not waiting a week for someone else either.

Comfortable enough that I do not have to queue behind the data team for basic questions. Joins, group by, subqueries, and window functions when I need a ranking or a running total. The four things I use it for most are sizing a problem before we agree it is worth solving, checking a stakeholder claim against reality, profiling data before a migration, and pulling baselines for a business case. The data profiling one has saved a project more than once. On a customer migration I ran counts on the fields the target system needed and found that about eighteen percent of records had no valid country code, which nobody had mentioned because the old system never enforced it. That turned into a data cleansing workstream we would otherwise have hit during cutover weekend. Where it goes beyond me is performance tuning or anything touching production write access, and I hand that to a data engineer without hesitation.