How It Works

Logs and Metrics go In, Incidents and Root Cause Come Out

Autonomous Incident Detection - how does it work

Step One - Ingest and Categorization

All you need to do is install install our log and metrics collectors - this takes less than 2 minutes.  No parsers, libraries, code changes or config needed. After that, our Machine Learning (ML) takes over.

Within minutes, our ML learns the structures of your logs,  and categorizes each event into a “dictionary” of unique event types. Categorization is crucial to being able to learn the patterns in your logs and metrics.

log and metrics collector setup

Zebrium machine learning

Step Two - Pattern and Anomaly Detection

In the first hour, it learns the normal patterns of every event and metric (the learning keeps getting better as it sees more data). 

When log or metric patterns change, our ML detects these as anomalies. But to separate signal from noise, it looks for hotspots of abnormally correlated anomalies across both metrics and logs. 

Step Three - Incidents and Alerts

The hotspots detected in step two are packaged into human readable incidents. Incidents make it easy for a user to clearly see the correlated set of anomalous log events and/or metrics.

Incident alerts are sent via Slack, email or webhook.

The entire process is completely autonomous - without the need for manual configuration,  user-defined thresholds or alert rules. 

2 - Zebrium Autonomous Incident

Getting started is free and easy

Spend just two minutes of your time and you'll be amazed at what we detect!