Sumo Logic Unveils Predictive Analytics and Machine Learning

Sumo Logic, the industry’s leading, secure and purpose-built cloud-based machine data analytics service, today announced the availability of Outlier Detection and Predictive Analytics capabilities that augment its rich and proven machine learning and Anomaly Detection engine with statistical analysis and projection models. The powerful capabilities make Sumo Logic’s service offering all-inclusive with built-in Pattern Detection, Anomaly Detection, Transaction Analytics, and now Outlier Detection and Predictive Analytics. The full spectrum capabilities empower DevOps, IT Operations and Compliance and Security teams to gain real-time visibility across thousands of data streams and detect and predict conditions that indicate an onset of performance, reliability or security issues within their apps and services.

“Outlier Detection represents an enormous opportunity to our teams to rapidly identify abnormal behavior in our KPIs, including response time, exception rates and network traffic,” said Dante Shareiff, Senior DevOps Engineer,TuneIn. “The ability to visualize and detect abnormal characteristics across multiple data sources eliminates the guesswork to identify the root cause of spikes and errors and helps expedite troubleshooting to keep our service running smoothly.”

Outlier Detection is powered by a unique algorithm that can analyze thousands of data streams with a single query, determine baselines and identify outliers in real-time. Purpose-built visualization highlights abnormal behaviors giving Operations and Security teams visibility into critical KPIs (Key Performance Indicators) and KRIs (Key Risk Indicators). Real-time alerts help teams react to and remediate critical issues as they are detected, such as a sudden rise in response time, unusual spike in network traffic or drop in request volume. Users can customize simple input parameters to manage sensitivity, baselines, direction and duration of change.

The Predictive Analytics capability extends and complements Outlier Detection by predicting future KPI violations and abnormal behaviors through a linear projection model. The ability to observe violations that may occur in the future, such as declining transaction volumes, rise in latency, and decrease in available application resources, helps DevOps, IT Ops, and Security teams address issues before they impact their business.

“Outlier Detection and Predictive Analytics help operators of mission-critical apps and services stay ahead of production issues,” said Bruno Kurtic, founding vice president of product and strategy at Sumo Logic. “Detecting an issue in behavior of a single host that indicates the onset of a larger problem or predicting a critical event a few minutes into the future can mean the difference between a system-wide outage and a minor operational procedure that goes unnoticed.”

Key Features and Benefits:
Outlier Detection
• Monitor multi-dimensional KPIs with dynamic thresholds
• Simple parameter tuning to customize KPI-specific analysis
• Purpose-built outlier visualization
• Real-time alerts on outliers

Predictive Analytics:
• Linear predictive analysis
• Real-time alerts for predicted future events

Availability
Both Outlier Detection and Predictive Analytics are currently available for Sumo Logic Enterprise Edition users for no additional fee.

For more information https://www.sumologic.com/press/2015-03-11/sumo-logic-unveils-outlier-detection-and-predictive-analytics-to-augment-machine-learning-and-anomaly-detection-capabilities/

Comments are closed.