# obs-unified

> Install, instrument, and read unified observability across every signal.

- [Welcome](/docs): Unified observability across traces, logs, AI calls, usage, replay, alerts, profiles, and analyses.
- [Getting started](/docs/getting-started): Start the unified collector and get multiple signal types flowing into one dashboard.
- [Installation](/docs/installation): Run obs-unified with the all-in-one Docker image, or install the editable local repo.
- [Examples](/docs/examples): Runnable examples, instrumentation guides, recipes, and migration paths.
- [SDKs](/docs/sdks): First-party SDKs for TypeScript, Go, and Rust on the backend; a browser SDK for usage and replay. They share one identity-propagation contract.
- [SDK API reference](/docs/sdk-reference): A compact cheat sheet for browser, TypeScript, Go, and Rust SDK APIs.
- [Instrumenting your app](/docs/instrumenting): React frontend + Node/Worker backend wired together so signals correlate end-to-end.
- [Python Flask instrumentation](/docs/instrument-python-flask): Send Flask traces and logs to obs-unified with OpenTelemetry and the interaction header.
- [What to expect](/docs/what-to-expect): How unified observability appears in the dashboard through the Connected rail.
- [Production operations](/docs/ops/production): Reverse proxy, Postgres tuning, storage retention, Kubernetes, and rollout checks.
- [How obs-unified compares](/docs/comparison): Vendor comparison with citations to Datadog, Sentry, PostHog, Honeycomb, New Relic, Grafana Cloud, SigNoz, Uptrace, and HyperDX.