TechnologyMachine Learning for Anomaly Detection

Corvil Intelligence Hub makes it easy for individuals with no data science expertise to use machine learning algorithms for anomaly detection. Corvil Intelligence Hub can detect anomalies based on deviations from learned baseline, changes in temporal patterns (such as phase, range and trend), predicted breach of key operational limits, and others. Corvil further enables correlation of detected anomaly alerts to reduce noise and provide more accurate, prioritized, actionable insight.

The algorithms are designed to be low touch and operate unsupervised, which enables users to apply them to the metrics of interest. Then Corvil’s real-time engine does the rest -- powering through high volume, velocity, time-series data generated by digital business transactions in real-time to:

  • Reveal Patterns: Automatically adjusting to changing behaviors, trends, and seasonality in the data
  • Identify Outliers: Pinpointing outliers and behaviors that differ significantly from past baselines or patterns.
  • Predict Conditions: Anticipates the future state of operational controls go beyond expected objectives


  • Business users can applied to metrics of interest
  • Insights derived from better data
  • Reduced alert fatigue
  • Faster time to value
  • Eliminated time consuming data preparation

Corvil Tops Info-Tech Research Network Monitoring Report Read the Report


Better Data, Better Insights

Algorithms leverage Corvil’s real-time, granular, precision-sequenced data to deliver better insights.

Broadly Applicable

Algorithms can be applied to any type of business, customer, or IT-centric metric.

Low Touch

Algorithms delivers reliable results without constant manual adjustments.

Real-Time Insights

Optimized for high volume, time-series data generated every nanosecond by digital business transactions.

Low Noise Alerts

Detected anomalies are automatically correlated, deduplicated, consolidated and filtered to minimize the noise.

Rapid Investigation

Explore details with multi-dimensional analysis and effective data visualizations