Before you can capture spans you must initialize the tracer. Initialization registers metadata about your application, environment, and any optional detections you want to run on traces.

​
Basic Initialization

initialize_tracer.py
from quotientai import QuotientAI

quotient = QuotientAI(api_key="your-quotient-api-key")

quotient.tracer.init(
    app_name="my-first-app",
    environment="dev",
)

​
Advanced: Adding Detections to Traces

For more advanced use cases, you can enable detections on your traces to automatically analyze spans for issues like hallucinations or document relevancy:
advanced_tracer.py
from quotientai import QuotientAI, DetectionType

quotient = QuotientAI(api_key="your-quotient-api-key")

quotient.tracer.init(
    app_name="my-first-app",
    environment="dev",
    detections=[DetectionType.HALLUCINATION, DetectionType.DOCUMENT_RELEVANCY],
)
The detections parameter accepts an array of detection types that will be automatically run on trace data when available.

​
Parameters

  • app_name (string): Human-friendly name displayed in the Quotient UI.
  • environment (string): Deployment environment (dev, staging, prod, etc.).
  • detections (array): Optional list of detections to run on traces. Currently supports HALLUCINATION and DOCUMENT_RELEVANCY.
  • instruments (array): Instrumentors that automatically attach to supported frameworks. See framework integrations.
  • tags (object, optional): Default metadata attached to every trace.

​
Tips

  • Initialize the tracer once at application startup—subsequent calls reuse the same configuration.
  • Combine tracer initialization with logger initialization so logs and traces share names and environments.
  • Pass instrumentors during initialization to automatically capture spans from supported frameworks.
Next: Record Traces