SQL.

Where SQL fits on a long engagement.

SQL is where the difference between a competent and an excellent specialist shows most directly in money. A query that scans a table instead of using an index is invisible in review, correct in output, and expensive every time it runs. Reading a query plan is a more valuable skill than knowing another framework.

It is also the fastest route to establishing what really happened. Before a bug is a code problem it is usually a data question, and a tester who can write the query finds the answer in minutes rather than filing a ticket for someone else to.

What an assigned team does with SQL.

SQL is where an assigned team quietly saves money. Query cost compounds daily, and nobody raises a ticket for a report that has always taken forty seconds. A specialist who has been on the product long enough to notice that it used to take four is the mechanism by which that gets fixed.

It also crosses the boundary between the people building the system and the people reporting on it, which is why the capacity is frequently split across two assignments on one agreement rather than one, with data science covering the analytical side.

What we use SQL for.

• Finding the cause rather than the symptom Querying production data directly to establish what state a record was actually in when a defect appeared.
• Performance work on a growing dataset Indexes and query shape revisited as volume changes, before the slow path becomes the outage.
• Reconciliation between systems Proving two systems agree, or finding exactly which rows they do not, which is the basis of most finance reporting.

How SQL capacity is assigned.

SQL is assumed rather than assigned for. Where it is the centre of the work — warehouse models, pipelines, reporting — that capacity is assigned under managed data services.