Why do AWS dashboards keep breaking — and is there a better way?

We’ve been running dashboards on AWS — mainly using Redshift with Power BI — and honestly, it's been rough. We’re dealing with slow load times, random breakages, and costs that are way higher than expected.

We’ve tried optimizing queries and resizing clusters, but it’s still messy. Is this just an AWS architecture issue, or are we doing something wrong?

Curious if others have moved away from AWS for dashboards or found a smarter way to handle this. Definitely open to hearing how others have solved it.