There are two cornerstones to building optimized and low maintenance queries in Power BI. The first is making sure queries have been designed to refresh at optimal speeds. Primary topics for performance tuning include: removing columns and rows, applied step configuration and order precedence, query folding, managing variables, and tables vs lists for transformations. The other important cornerstone is intelligently identifying and handling errors as they occur during scheduled refreshes. Primary topics for error handling include: replacing vs removing errors, collecting error metadata, and configuring the refresh to either fail or refresh successfully when errors occur.