It is important to note that Rapid7 InsightCloudSec's features are not 100% precise, but I find about 70% of the time it is satisfactory. I would like to suggest that you improve it to be more precise, ideally making it 100% if possible. Some cases in Rapid7 InsightCloudSec indicate that the log is not enough, as they mostly just generate alerts, and the synchronization between data connectors is often problematic, particularly in terms of not being in sync always, especially between the AD and Rapid7 alerts, which generates numerous false positives. Additionally, the traditional rules should be updated, as this is a main point worth mentioning since we spend a lot of time fine-tuning these traditional rules. I suggest improving the legacy detection rules. If there are any authentication cases, such as impossible travel activity where a user has their SharePoint hosted in a different location, Rapid7 can often trigger alerts, creating confusion as we cannot fine-tune it properly. Another issue is with honeypot access. We sometimes lack necessary logs because Defender's advanced threat protection scanning gets detected as honeypot activity by Rapid7, leading to annoying and noisy alerts that we need to constantly close. If you can improve the traditional detection rules to reflect current detection rules, it would make it significantly easier for us to manage, as we constantly need to check legacy rules to update or possibly turn them off. Updating the legacy rules should be a priority.
