Overview

​

Introduction

Cased’s anomaly detection system helps you discover, understand, and respond to issues that occur after deployments. By connecting to your existing monitoring tools, Cased automatically identifies problems and correlates them with recent changes in your system.

​

Key Features

​

1. Automatic Detection

• Continuous monitoring of your applications
• Integration with multiple data sources (Sentry, Datadog, AWS CloudWatch)
• Real-time analysis of error patterns and performance metrics
• Configurable thresholds for different types of anomalies

​

2. Deployment Correlation

• Automatic linking of issues to recent deployments
• Intelligent code analysis to identify probable causes
• Timeline visualization of issues relative to deployments
• Confidence scoring for correlations

​

3. Smart Notifications

• Real-time alerts for critical issues
• Configurable notification thresholds
• Integration with your existing communication tools
• Context-rich notifications with direct links to relevant information

​

4. Response Automation

• Auto-rollback suggestions for serious issues
• PagerDuty integration for incident management
• Quick actions for common response patterns
• Team collaboration tools

​

Getting Started

​

1. Connect Your Tools

Configure your monitoring sources in the Cased dashboard:

​

2. Set Up Thresholds

Define what constitutes an anomaly for your system:

Currently high volume anomalies are detected by default with the following thresholds:

Error events: 10 counts

Please reach out to support to customize thresholds.

​

3. Configure Notifications

Choose how and when you want to be notified

• High volume anomalies

• Deployment-related anomalies

​

4. Notifications examples

• High volume anomaly notification

• Deployment-related anomaly notification

• Deployment-related metrics spike notification

• Auto-rollback suggestion