Overview

LogicMonitor’s March 31 Product Power Hour highlighted how AI-driven investigations are becoming more precise, explainable, and operationally useful for technical teams working through incidents at speed. The session focused on recent enhancements to Edwin AI, including problem management, topology-based correlation, deeper log-driven root cause analysis, and change-aware investigations that help teams move from alert to action faster.

A major theme throughout the conversation was trust. Attendees wanted to understand not just what Edwin recommends, but why it made that recommendation, what evidence it used, and how that information can be pushed into the tools their teams already rely on. The product team addressed that head-on by showing how Edwin surfaces logs, change requests, topology context, and knowledge sources directly inside the investigation experience.

Key Highlights

🌟Edwin AI now provides richer AI investigations with natural language summaries, root cause analysis, and actionable remediation guidance based on alert, log, and change data.

🌟A new problem management experience helps teams identify recurring issues across the last 90 days so they can focus on deeper root cause resolution instead of repeatedly treating symptoms.

🌟Topology-based correlation is helping Edwin group alerts more intelligently by understanding device relationships and dependency patterns.

🌟The team shared that ServiceNow change integration is available today, with Jira planned next.

🌟Customers showed strong interest in extending Edwin with additional data sources, especially Catchpoint, SOP and knowledge content, and metrics-driven reporting.

🌟The discussion reinforced that users want Edwin to support both workflows: deep investigation in the Edwin UI and pushing relevant context into ITSM systems.

Q&A

Q: Is build of the topo maps dependent on monitoring protocols used? For example, SSH versus SNMP to monitor network devices?
A: Yes. How we pull information from LogicMonitor depends on their dedicated topology sources, which use a mix of four or five different protocols to build the maps. It depends on the type of devices you are monitoring in LogicMonitor, and that dictates how the maps are built. We also are not limited to only network and network maps in the Element Vision side. If you have VMs, you will still get a map or connections that we can pull through to Edwin soon as well.

Q: Can you use topology maps from other products such as NetBrain?
A: Not right now. We are currently experimenting with Dynatrace, which has an MCP available that lets us fetch topology information. In theory, if NetBrain or another third party has an equivalent way for us to fetch topology information, then yes, we would be able to layer that into what we have. The key requirement is access to topology information in a format where we can determine how systems are connected.

Q: What sources can you tap into for SOP-type material?
A: Right now we are connected to ServiceNow, so you can set up the LMDX integration to allow us to read those articles. Next on the list are Confluence articles and SharePoint via MCP. If there is another data source you are interested in, please share it, because we are able to start adding those integrations. We can also do a public knowledge search, so if you want to tap into vendor documentation or other publicly available web content, you can query that from the agent too.

Q: Is the change integration available today?
A: Yes, it is with ServiceNow. If ServiceNow is your ITSM, that is available today. We are working on Jira next for changes.

Q: It seems like there used to be an available Edwin AI Python SDK integration under Settings. I do not see it now. Am I misremembering it being available?
A: It is available. We might just have to pass you the files for the SDK, because I do not think it is available to download anywhere. It should include how you send through events if you want to set up a custom integration. The other option is to use the webhook if you want to send events or information into Edwin, and it sends to that same endpoint as long as you include the required fields.

Q: Is there an article for the webhook approach too?
A: Yes. We do have documentation, although it probably needs to be updated because it does not clearly list the endpoint, which is likely the most helpful part. Right now it mostly covers the required fields, but we can make sure it gets updated so the endpoint is more obvious.

Q: Could you tell us what’s behind the AI at LogicMonitor? What technologies, approaches, or architecture are used to build and operate this solution?
A: It varies depending on the use case. Forecasting, for example, is best handled with a regression algorithm based on historical data. Edwin AI leverages both ML models and agentic frameworks to correlate alerts and fetch information from third-party sources. We try to match the type of AI we use to the particular use case we are solving for.

Q: Why was the AI in LogicMonitor named Edwin? Is there a specific meaning or story behind that name?
A: It was named after Edwin Hubble. The idea is that the Hubble telescope helped discover and reveal more of what is out there in space, and Edwin AI is meant to do something similar for your observability systems by combining information and giving you insights into your data.

Customer Call-outs

⭐“Pushing the data into ITSM is necessary for sure. The additional functionality added to Edwin makes it compelling for SA's to use the Edwin UI.”

What's Next

⚡Product Power Hour

Smarter Alerting that Drives Faster Action

April 23rd | 1pm CT

The next Product Power Hour will focus on alerts, including alert tuning, management, routing, and upcoming alert-related features. This session is a strong follow-up for teams looking to improve signal quality and operational response workflows.

👉Register here today

🪵Logs for Lunch

More Value and Less Worry with Log Partitions

April 8 | 12pm CT

The next Logs for Lunch session will cover log partitions and will take place at 12:00 p.m. Central Time. This is a useful session for practitioners looking to deepen their log management strategy.

👉 Register here today

🍽️ User Group Dinners

Join us as we head back on the road! Connect in person with other LM users in your city over dinner and real talk. Share wins, swap stories, and grow your network.

San Francisco/Bay Area - April 28
Chicago - May 27

Resources and Review

🏅Want to upskill? Earn your Edwin AI badge!

💻 Review the slide deck here

📹Watch the recording here or directly on the event page.