A report could display a filtered page, yet exporting the same selection took much longer. It already used a streaming workbook writer, making batch size or writer configuration plausible places to investigate.

Tracing the provider exposed repeated work upstream. Every output batch entered the …

A production build reported several megabytes of JavaScript, while the application entry compressed to a much smaller file after code splitting. Neither number answered the operational question: how much code did an unauthenticated user actually need before the login screen became usable?

A …

A paginated operations screen took 27.49 seconds to return ten kilobytes of data. The small response ruled out payload transfer as the main problem, and the browser timing alone could not explain whether the delay came from rendering, application code, or the database.

The useful clue was that one …

An order-status dashboard worked in the test environment but showed zero for every inbound, outbound, and workflow counter in production. The page list contained many orders, and read-only database checks confirmed recent activity. Zero was not the business result; it was the value left behind when …

A wave planning list in a WMS application took about six seconds to return one page of results. This note records the investigation path and the evidence behind the final optimization direction.

Identifiers and environment details have been anonymized. The SQL shape, measurements, and decisions are …