CrunchMetrics Anomaly Detection Software

Key Features

• Real-time Anomaly Detection

  Monitors data streams to identify abnormal patterns and alert stakeholders.

• Self-Learning Algorithms

  Uses ML and deep learning to evolve its understanding of normal behavior as the database grows.

• KPI Correlation

  Supports linking multiple KPIs and establishing parent-child relationships to help identify the root cause of an incident.

• Anomaly Scoring

  Assigns a statistical score to each deviation to indicate the severity of the anomaly.

• Event-Based Learning Control

  Allows users to exclude specific data points or time frames from retraining to prevent known events from influencing the model.

Use Cases

• Payment Success Monitoring

  Using deep learning to identify dips in payment success rates for financial service providers.

• Pricing Error Detection

  Monitoring contextual anomalies in e-commerce to identify pricing errors or mis-promotions.

• Network Performance Tracking

  Identifying early indicators of network degradation for telecom operators.

• Time-Series Trend Analysis

  Detecting structural breaks and shifts in data while controlling for seasonality like holidays or days of the week.

Integrations

• API
• Web services

FAQ

What is CrunchMetrics used for?

It is used to automatically identify abnormal behavior or deviations in business data, such as payment failures or network drops, in real time.

Does it work with existing BI tools?

It is designed to integrate with existing data warehouses and BI systems through APIs and web services.

Can it handle seasonal data changes?

The algorithm is designed to control for effects like the day of the week, holidays, and other special events to identify true anomalies.

Is there a free trial?

The evidence indicates that a free trial option is available for users to test the platform.