metrics · traces · logs · rum
Unified signal stream
Metrics, distributed traces, structured logs, and real-user monitoring flow into one timeline. Correlate a spike in p99 latency with the deploy that caused it — in two clicks.
Features
Every signal. One coherent light.
Lumen treats observability as infrastructure: invisible when healthy, instantly legible when something shifts.
live · <10s lag · embeddable
Ambient dashboards
Boards that update in near-real-time without a refresh. Surface the KPIs that matter on wall screens, in Slack, or embedded in your incident command channel.
adaptive · ML baseline · dedup
Intelligent alerting
Adaptive thresholds that learn your traffic shape — so Monday morning scale-ups stop paging the on-call team. Alert on anomaly, not on threshold guesses.
OTel · flame graph · service map
Deep trace explorer
Flame graphs and service maps rendered from OpenTelemetry spans. Drill from a slow HTTP endpoint down to the exact SQL statement adding 400ms.
auto-correlate · root cause · RCA
Incident timeline
Every alert, deploy, feature flag, and anomaly on a single timeline. Stop reconstructing incidents from Slack. The evidence is already correlated.
200+ integrations · OTel native
Integrations without friction
One-line agent, zero-config Kubernetes operator, or OpenTelemetry Collector — connect your stack in minutes. 200+ integrations, all maintained by Lumen.
99.99%
Ingestion uptime over the last 12 months
8.4B
Spans ingested every day
<10s
Signal-to-dashboard lag at p95
200+
Maintained integrations
How correlation works
From signal to root cause in one timeline.
Every metric anomaly, trace error, log pattern, and deploy event flows into Lumen's correlation engine. It matches them by timestamp, service boundary, and causal graph — surfacing the most likely root cause before you finish typing in the incident channel. No context switching between APM, logging, and deploy dashboards. The evidence is already assembled.
$ lumen incident --latest incident: INC-2841 (auto-detected · 14s ago) signal: p99 latency spike on /api/checkout (+340ms) trace: db.query "SELECT * FROM orders" → 380ms avg correl: deploy #1047 (schema migration) · 2m before spike action: rollback deploy #1047? [y/N]
Your whole system, in one light.
Stop guessing. Start seeing.
Connect your first integration in under five minutes. Your first 5 GB of signals are free — no card, no commitment.