Splunk Inc. today introduced a set of new artificial intelligence features for its widely used data analytics products that aim to help companies better identify issues with their technology infrastructure, as well as preempt future outages.

The first offering that the company has updated is Splunk IT Service Intelligence. The new iteration of the tool features machine learning models, that, according to the provider, can spot signs of an imminent infrastructure or application outage.

Other companies are applying AI in similar ways. General Electric Co., for example, offers a platform called Predix that enables its industrial customers to predict equipment malfunctions ahead of time. Splunk IT Service Intelligence takes the core concept a step further by giving administrators insight into exactly how an imminent outage would impact operations. It also automatically prioritizes incidents based on severity.

The latter feature can come quite handy given the massive amounts of operational data generated in a typical enterprise. By filtering out everyday “event noise,” as Splunk calls it, the tool should enable administrators to identify the most urgent issues faster. In the same spirit, the company also enhanced Splunk IT Service Intelligence’s existing algorithms for detecting security problems.

The update debuted alongside a new release of the company’s Machine Learning Toolkit, which is likewise offered as an add-on for its flagship Splunk Enterprise analytics platform. The biggest addition is a visual interface for monitoring AI projects.

The Machine Learning Toolkit lets companies build custom models to analyze the data they process with Splunk Enterprise. According to Splunk, engineers can now visually track key information about a project such as how many operational anomalies were detected in a given dataset. Users also have access to new AI training algorithms designed to help hone a machine learning model’s pattern recognition capabilities.

Splunk said companies can apply Machine Learning Toolkit’s features to process a variety of different data types. Hyatt Hotels Corp., for example, has built a model to monitor WiFi usage at its locations for signs of traffic drops that may indicate users are experiencing connectivity issues.

Topping off today’s announcement is an update to Splunk Enterprise and Splunk Cloud, the managed version of the platform. The main highlight is a pair of new data integrations with popular open-source technologies. The first connector is for Kubernetes and Docker while the other is designed to work with Apache Kafka, a tool used by large enterprises to shuffle information between systems.

Image: Splunk