Signal
Start monitoring
Now shipping OpenTelemetry-native ingestion

Every signal that matters, in one dense view.

Signal replaces the dashboard-of-dashboards with a single operator surface: live service tables, sparklines, SLO burn-rate alerts and incident correlation — built for people who read numbers, not gradients.

Start free trial Watch a live view
Services tracked
128,406
Alerts triaged / wk
42,318
Median MTTR
4m 12s
app.signal.dev/services
Services · production
6 healthy 1 degraded
Service Status p95 Err% 30d
api-gateway up 118ms 0.02%
checkout-svc degraded 612ms 2.41%
postgres-primary up 4.1ms 0.00%
event-worker-03 up 89ms 0.05%
redis-cache-a up 0.9ms 0.00%
Auto-refreshing every 10s Open full view
Incidents auto-correlated
91%
of pages linked to a root-cause change
Alert noise reduced
63%
fewer duplicate pages per on-call shift
Operator teams
1,940
running Signal in production
Platform uptime SLA
99.99%
trailing 12 months, published status page

Built for operators, not for demos.

Six capabilities that replace a stack of point tools — correlation, SLOs, tracing, paging, anomaly detection and audit exports, all reading from the same event stream.

Incident correlation

Groups related alerts, deploys and topology changes into one incident timeline so on-call sees cause before symptom.

SLO burn-rate alerting

Multi-window burn-rate rules page before error budgets are exhausted, not after the postmortem.

Distributed tracing

Full-fidelity spans across services with tail-based sampling, pinned to the exact row that triggered the trace.

On-call scheduling & paging

Rotations, overrides and escalation policies live next to the alerts they page for — no context switch to a separate tool.

Anomaly detection

Seasonal baselines per metric catch slow leaks and traffic-shape drift that static thresholds miss.

Audit-ready exports

Every state change, ack and resolution is timestamped and exportable for SOC 2 and internal postmortem review.

One dense view, three lenses.

Switch between services, infrastructure and open alerts without leaving the row you were reading. Density is adjustable; the data underneath never is.

Service Status p50 p95 Err% 30d uptime Trend Owner
api-gateway operational 42ms 118ms 0.02% 99.98% RK
checkout-svc degraded 210ms 612ms 2.41% 99.61% MT
payments-core down — — 100% 98.90% JL
postgres-primary operational 1.2ms 4.1ms 0.00% 99.99% RK
event-worker-03 operational 31ms 89ms 0.05% 99.97% SN
redis-cache-a operational 0.4ms 0.9ms 0.00% 100.00% MT
search-indexer operational 55ms 140ms 0.08% 99.94% JL
billing-cron operational 890ms 2.1s 0.00% 99.99% SN
Host Region CPU Mem Disk Status
ip-10-0-4-112 us-east-1a 64% 71% 38% nominal
ip-10-0-4-118 us-east-1a 91% 88% 42% hot
ip-10-0-6-004 us-west-2b 33% 55% 60% nominal
ip-10-0-6-011 us-west-2b 28% 49% 61% nominal
ip-10-0-9-201 eu-central-1 — — — unreachable
Alert Severity Service Age Assignee Status
payments-core error rate > 5% sev1 payments-core 00:11:42 J. Liu paged
checkout-svc p95 latency budget burn sev2 checkout-svc 00:38:05 M. Torres acked
ip-10-0-4-118 sustained CPU > 90% sev3 infra 01:52:19 Unassigned open
eu-central-1 node unreachable sev1 infra 00:04:57 R. Kaur paged

Recent deploys

  • 08:41
    checkout-svc v2.14.1 — no regressions
  • 07:02
    api-gateway v5.3.0 — rolled to 100%
  • 05:18
    payments-core v3.9.2 — correlated with sev1
  • 02:47
    event-worker-03 v1.1.8 — no regressions

From threshold breach to resolved, tracked.

Every alert moves through the same lifecycle. Signal keeps the audit trail so postmortems write themselves.

  • Detected
    Metric crosses a static or anomaly threshold
  • Correlated
    Grouped with related alerts and recent deploys
  • Paged
    Routed to the on-call schedule with escalation
  • Acknowledged
    Owner claims the incident, timer starts
  • Resolved
    State closed, timeline exported for review
Preview an exported incident record
{
  "incident_id": "INC-20461",
  "service": "payments-core",
  "severity": "sev1",
  "detected_at": "2026-09-21T08:04:12Z",
  "acknowledged_by": "[email protected]",
  "resolved_at": "2026-09-21T08:19:07Z",
  "mttr_seconds": 895
}

Operators run on Signal.

"We cut our median time-to-acknowledge from 9 minutes to under 90 seconds after moving on-call off five tools and onto Signal's single table view."

Portrait of Dana Aoki
Dana Aoki
Director of Site Reliability, Northwind Freight

"Alert fatigue dropped fast once duplicate pages got correlated automatically."

Portrait of Priya Kapoor
Priya Kapoor
Staff Engineer, Ledgerline Payments

"The SLO burn alerts caught a slow leak two days before it would've paged at 3am."

Portrait of Tomas Ortega
Tomas Ortega
Platform Lead, Harborline Logistics

Priced per monitored service, not per seat.

Add every operator to the on-call rotation without paying more for eyeballs on the dashboard.

Team
$49 /mo

Up to 25 monitored services

  • Service & infrastructure tables
  • On-call scheduling for 10 operators
  • 7-day alert history
  • SLO burn-rate alerting
Start free trial
Most popular · Business
$199 /mo

Up to 250 monitored services

  • Everything in Team
  • SLO burn-rate alerting
  • Incident correlation & distributed tracing
  • 90-day alert history & audit export
Start free trial
Enterprise
Custom

Unlimited services, dedicated ingest

  • Everything in Business
  • SSO, SCIM & audit-log streaming
  • Unlimited alert & trace retention
  • Dedicated solutions engineer
Talk to sales
What counts as a monitored service?
Any process, database, cache or queue that Signal ingests health, latency or error metrics for — regardless of host count behind it.
Can I self-host the ingest pipeline?
Business and Enterprise plans support a self-hosted collector that forwards to Signal's managed storage and alerting layer.
Do you support OpenTelemetry natively?
Yes — OTLP over gRPC and HTTP are first-class ingestion paths for metrics, traces and logs, no custom agent required.
What happens if I exceed my plan's service limit?
Signal keeps monitoring every service and emails your billing owner; nothing is dropped or rate-limited mid-incident.

Stop scrolling through noise. Start seeing signal.

Connect your first service in under five minutes. No credit card required for the 14-day trial.

Start free trial Read the docs