LogicMonitor Inc., maker of an observability platform that companies use to find and fix technical issues in their technology infrastructure, today announced a series of enhancements that are aimed at enhancing automation and improving insights for reliability engineers.

The release provides greater visibility into artificial intelligence workloads and applications and upgrades LogicMonitor’s Edwin AI generative AI agent. AI workload monitoring is being improved with expanded support for Amazon Web Services Inc.’s Q Business gen AI assistant and Nvidia Corp. graphic processing units to allow information technology teams to monitor and optimize AI-driven applications from a single control point.

AI cost monitoring

“If customers are building AI on top of the AWS platform with Bedrock, we are integrated into the stack so we can monitor the workload and report on token consumption and the cost of tokens,” said Karthik Sj, general manager of AI.

New support for Amazon’s managed Elastic Kubernetes Service and Microsoft Corp.’s Azure Kubernetes Service provides improved visibility into AI workloads deployed in cloud-based software containers, which are lightweight, portable units that package applications and dependencies together to ensure consistency across different computing environments. That enhances AI workload reliability, the company said.

“We have comprehensive Kubernetes monitoring today including OpenShift, Google Kubernetes Engine and Rancher,” which is an open-source enterprise Kubernetes manager, said Sarah Terry, vice president of product management at LogicMonitor. “What we found is ITOps teams don’t always have direct access to clusters. We’re introducing integration through the cloud provider API [application program interface] to provide higher-level visibility into the cluster without requiring cluster access.”

New cost optimization dashboards combine cost visibility and recommendations to help IT teams manage compute-intensive AI workloads more efficiently by balancing performance, cost and sustainability.

AI-powered alert filtering

Edwin AI, the generative AI assistant LogicMonitor introduced last June, is claimed to deliver up to a 90% reduction in alert noise reduction by autonomously automating troubleshooting, prioritizing critical alerts and accelerating resolution workflows.

The agent now has AI-powered alert filtering and prioritization to surface the most critical incidents first. Instance-level metadata correlation and AI-powered recommendations cite relevant past incidents to speed troubleshooting. Edwin AI can now also connect to PagerDuty Inc.’s incident management platform, Dynatrace Inc.’s observability suite, ConnectWise LLC’s service management platform and other IT operations tools.

Enhancements to LogicMonitor’s embedded log analysis capabilities now provide instant access to relevant log data and AI-powered log correlation without the need for complex query languages, the company said.

“We can use machine learning models to correlate attributes across different devices,” Sj said. “Gen AI can be used to summarize an alert to make it more accessible.” LogicMonitor currently manages gen AI prompts but intends to expose them to customers in the future.

Image: SiliconANGLE/Ideogram