See everything, own your data: the Giant Swarm Observability Platform

Most platform teams have an observability stack. Few would call it finished. You start with metrics, bolt on logging, realize you need traces, wire up alerting — and twelve months later, two engineers spend their days keeping the monitoring alive instead of using it. That's the observability assembly tax, and your team deserves better than paying it.

That's why we built the Giant Swarm Observability Platform. Metrics, logs, and traces in one curated, integrated platform: 100% open source, OTLP-native, running in your environment, production-ready from day one.

What you actually get

Three things that matter to your team and your budget:

1. See everything. Metrics, logs, and traces correlated in one place. When an alert fires, the path from metric anomaly to relevant logs to the exact trace that shows the bottleneck is one click, not three separated tools.
   
   
2. Own your data. The platform runs in your environment — your data never leaves your infrastructure, and every component is open source. It connects to your existing OIDC provider, so access management stays where it belongs. If you ever move on, you keep everything: the tools, the data, the dashboards. Zero lock-in.
   
   
3. Costs that scale transparently. No per-GB ingestion fees. No per-seat licensing that grows with your headcount. No surprise bills when traffic spikes during a product launch. Your costs scale with the infrastructure you actually run — not with opaque vendor pricing models. That means you can plan capacity and budget with confidence, because the relationship between usage and cost is something you control, not something a vendor decides.

What's under the hood

The Observability Platform is a fully integrated capability in Giant Swarm's Curated Platform Engineering Stack. Metrics, logs, and traces — the three pillars of observability — in one platform, production-ready from day one.

The technology is 100% open source, built on the LGTM stack — Loki for logs, Grafana for visualization, Tempo for traces, Mimir for metrics — plus Alertmanager for notifications and Alloy as the OpenTelemetry collector. The platform natively supports OTLP, so your teams can instrument with any OpenTelemetry-compatible SDK and send data straight in — no adapter layers, no vendor-specific agents.

You could run the LGTM stack yourself, but that's where the assembly tax starts. What makes this a platform is everything on top: correlated data across all three pillars, proven configurations from 150+ production clusters, multi-tenancy by default, GitOps-ready dashboard management, and self-service alert rules, routing, and silences.

This is the observability stack we run ourselves — across our entire customer fleet, 24/7. Every configuration shipped with the platform comes from real operational experience solving problems like these:

• Incidents take too long to resolve — correlated metrics, logs, and traces cut the path from alert to root cause, reducing mean time to repair (MTTR)
  
• Nobody knows what's driving cost — unified resource visibility makes infrastructure cost drivers obvious, not buried in spreadsheets
  
• Alert fatigue drowns real signals — multi-dimensional SLO-based alerting replaces noisy threshold alerts with meaningful signals
  
• Teams can't self-serve — self-service access to alerting, dashboards, and data management scales team autonomy without bottlenecking on a central ops team
  
• Dashboards are inconsistent or missing — curated, pre-built dashboards provide a consistent starting point that teams can customize for their services
• Decisions rely on gut feeling — a unified, single source of truth for operational data enables evidence-based decision making across teams

That experience is what you get out of the box; proven defaults, curated configurations, and system data collection that works before you touch a single setting.

Available today

The Observability Platform is available now — as a standalone capability or fully integrated into Giant Swarm's curated platform stack. If your team is paying the observability assembly tax — whether in engineering hours or vendor bills that grow faster than your business — let's talk.