Smart alerts.
Zero noise.
Threshold, anomaly, burn-rate, absence, and composite rules: routed to Slack, PagerDuty, OpsGenie, Microsoft Teams, email, or webhooks. Silences, alert history, and AI-suggested rules keep your on-call signal high and your pages low.
Deep Dive
Core Capabilities & Architecture
Explore how Alerting instruments, processes, and correlates high-cardinality telemetry at scale.
Five rule types for every kind of signal
Static thresholds for the obvious, Z-score anomaly detection for the subtle, SLO burn-rate for reliability budgets, absence for dead pipelines, and composite rules to combine them with AND/OR logic.
- Threshold rules with GT / LT / EQ / NEQ conditions
- Z-score anomaly alerts that learn each metric’s baseline
- Multi-window SLO burn-rate alerts for fast and slow burns
- Absence and composite (AND/OR) rules in one engine
Notify the right people, everywhere they work
Route every alert to Slack, PagerDuty, OpsGenie, Microsoft Teams, email, or any webhook. Each notification carries trace IDs, log context, and a metric timeline so responders start with answers, not questions.
- Slack, PagerDuty, OpsGenie, Microsoft Teams, email, and webhooks
- Test any channel before you trust it with a 3 a.m. page
- Rich context: trace ID, correlated logs, metric timeline
- Per-rule severity: CRITICAL, HIGH, MEDIUM, LOW
Cut the pages that don’t matter
Silence alerts during deploys and maintenance with time-bound, label-based windows. Cooldowns prevent flapping, and AI suggests rules from your own queries and metrics.
- Label-based silences for deploys and maintenance windows
- Cooldown windows that stop alert storms before they start
- Full alert history for every fire and resolution
- LLM-suggested alert rules from a query or metric
Built for every team that cares about reliability
One platform, tailored to how your team actually works.
Fewer, better pages
Anomaly and composite rules catch real problems without paging on every blip.
Protect the error budget
Burn-rate alerts fire before an SLO is breached, not after the incident.
Alerts as code
Standardize rules and routing across teams, with silences for planned work.
Five rule types, six notification channels, one engine
Threshold, anomaly, burn-rate, absence, and composite rules evaluate against the same ClickHouse telemetry your dashboards use, no separate alerting datastore to drift out of sync. Every rule routes to Slack, PagerDuty, OpsGenie, Microsoft Teams, email, or webhooks with full context attached.
Also in the platform
Guides
Start smarter alerting in 5 minutes
Your existing OpenTelemetry instrumentation works unchanged, so there is nothing to rewrite and nothing to migrate.