The Problem

SREs need to quickly identify and respond to system anomalies.

Technical Challenge

Metric spikes often lack context about similar past incidents and their resolutions.

Our Solution

Clustered anomaly detection with salience-ranked recall across historical incidents.

Example Usage

attend cluster "latency spike" --timeframe 7d --metrics cpu,mem,latency

Key Benefits

• Instant recall of similar past incidents
• Automated correlation of metrics and logs
• Proactive anomaly pattern detection
• Contextual playbook recommendations