Zero-Config OpenTelemetry Setup: From Zero to Production in 5 Minutes

The Pain of Traditional Instrumentation

Setting up traditional observability pipelines often involves days of configuring collectors, exporters, and proprietary agents. Teams spend more time debugging their monitoring setup than actually using it. The irony of modern DevOps: the tools designed to tell you when your application is broken often need an application engineer just to keep them running.

ObservabilityOS takes a different approach. Our zero-config SDK bundles everything you need into a single dependency: a structured logger, an OTel trace exporter, a metrics collector, a PII scrubber, and an async ingestion queue. No collector daemon, no YAML pipelines, no exporter routing.

One Command Setup

Install the SDK in any Node.js or TypeScript project. The SDK auto-patches console.log, captures unhandled exceptions, begins streaming structured telemetry to the ObservabilityOS ingestion endpoint, and enables PII scrubbing by default.

npm install @observability-os/sdk

# In your main entry point — before any other imports:
import '@observability-os/sdk/register';

# Environment variables:
OBSERVABILITY_API_KEY=your_key_here
SERVICE_NAME=checkout-api
NODE_ENV=production

What Happens Under the Hood

On initialization, the SDK creates an in-memory ring buffer for high-throughput batching. Logs are stored instantly in memory and flushed asynchronously in the background every 2 seconds or when the buffer reaches 100 entries — whichever comes first. The non-blocking queue ensures zero added latency to your application.

The built-in scrubber automatically redacts credentials, JWT tokens, credit card numbers, and custom regex patterns before any data leaves your process. Your cloud indexer only ever receives sanitized payloads. Within seconds of your first log, the dashboard displays your service health, error rates, and latency. No configuration required.