Adaptive Alert Detection

Innovation Overview

Adaptive Alert Detection uses a sophisticated machine learning (ML) algorithm to deliver more intelligent, accurate alerts. It dynamically adjusts to real-time network conditions, reducing false positives and minimizing flapping alerts—all without the need for manual configuration or fine-tuning. Most customer environments will benefit from the system's default medium sensitivity setting, which delivers an out-of-the-box value with no additional setup required.

Feature Highlights:

• Automatic Aggregation: Works right out of the box with medium sensitivity without fine-tuning the alert occurrence and threshold configuration.

• Simplified Alert Rules: Customers can select the alert criteria and adjust sensitivity levels without worrying about how often the requirements must be met before triggering the alert. Instead, ThousandEyes detects anomalies and issues automatically.

• Reduced Flapping Alerts: Differentiates between triggering and clearing conditions, minimizing unnecessary alert noise.

Customer Benefits:

• No More Alert Fatigue: Fewer false positives and better detection of genuine issues.

• Zero Configuration Hassle: Adaptive aggregation happens automatically, allowing teams to focus on fixing problems rather than configuring alerts.

• More Intelligent Context-aware Alerts: The system learns from network patterns to deliver actionable alerts, helping teams resolve incidents faster.

Here is a quick tutorial on setting up Adaptive Alert Detection:

1. Create a new Alert rule using Adaptive Alert Detection

2. Switch between Adaptive alerting and Manual alerting

3. Dynamic Baseline condition is inherited in Adaptive Alerting