When it comes to the improvement of LogicMonitor, I think there are a few points that can be improved. The first one is alert tuning, which takes time. It requires effort when trying to understand it for the first time. The defaults do not always match our workload patterns, so I have to adjust the thresholds to reduce noise and avoid alert fatigue. While the dashboards are solid, I sometimes wish that the UI was a bit more intuitive when drilling down quickly during an incident. There are many options and finding the exact view where I can identify the exact problem takes a few extra clicks. When an alert comes and I click on a LogicMonitor alert, it takes time to understand what the alert actually is and to go through the data points. The alert page specifically could be better. The alert tuning part can also be made more simple. The first area that could be better is alert clarity and routing. Sometimes alerts do not include enough immediate context, so I still have to spend a few minutes correlating data across views. Adding more actionable details directly in the alert would make the response even faster. LogicMonitor sometimes gives false alerts as well. For example, if an EC2 instance is down, it will not determine whether the EC2 instance has been deliberately turned off or if it is actually not responding. At that time, it will give false alerts. The clearing of alerts is also an issue. Once an issue is fixed, the alert should be cleared, but it takes a little time for that alert to be cleared. Another improvement that would be helpful is simpler customization for complex dashboards. It is powerful, but building highly tailored dashboards, especially across multiple environments, can feel heavy and time-consuming. I would also appreciate a stronger out-of-the-box AWS correlation, such as automatically grouping related issues across EC2, EBS, and ALBs in a way that reads as a single incident story. This would reduce the mental overhead during outages. Grouping incidents together, such as all the EC2 alerts, all the EBS alerts, or all the load balancer alerts would be beneficial. Overall, none of these are blockers, just some improving areas. There could be smarter anomaly detection out of the box that can catch unusual but important behavior without manual tuning of every threshold. Better tagging and dynamic grouping for EC2 instances would also be helpful. Cleaner alert de-duplication so a single underlying issue does not generate multiple redundant alerts would improve the system. More guided root cause workflows would be beneficial, such as providing the most likely causes based on correlated metrics. Faster search navigation across devices, dashboards, and alerts during incidents would also improve the platform.
