docs: Add MLflow tracing example for voice agents

[B-Step62]

Add mlflow_trace.py example that demonstrates how to send OpenTelemetry traces from LiveKit Agents to MLflow's tracking server, following the structure of other tracing providers. MLflow is an open-source LLMOps platform with 30MM monthly download.

Doc on MLflow side: https://mlflow.org/docs/latest/genai/tracing/integrations/listing/livekit/

[CLAassistant]

Thank you for your submission! We really appreciate it. Like many open source projects, we ask that you sign our Contributor License Agreement before we can accept your contribution.
_{You have signed the CLA already but the status is still pending? Let us recheck it.}

[devin-ai-integration]

Devin Review found 1 potential issue.

View 3 additional findings in Devin Review.

[devin-ai-integration]

🟡 experiment_id parameter silently ignored when OTEL_EXPORTER_OTLP_HEADERS env var is set

When a user calls setup_mlflow(experiment_id="5") but the OTEL_EXPORTER_OTLP_HEADERS environment variable is already set, the explicitly passed experiment_id parameter is silently ignored.

Root Cause

On line 58, experiment_id is resolved from the parameter (or defaults to "0"). But on lines 62-63, the code does:

os.environ["OTEL_EXPORTER_OTLP_HEADERS"] = os.getenv(
    "OTEL_EXPORTER_OTLP_HEADERS", f"x-mlflow-experiment-id={experiment_id}"
)

This means if OTEL_EXPORTER_OTLP_HEADERS is already set in the environment, the env var value takes precedence and the experiment_id parameter is discarded. This is inconsistent with how endpoint works on line 57, where the parameter takes precedence over the env var via endpoint = endpoint or os.getenv(...).

Impact: A user explicitly passing experiment_id would expect it to be used, but traces could be sent to the wrong MLflow experiment if OTEL_EXPORTER_OTLP_HEADERS is already set in the environment.

Suggested change

	os.environ["OTEL_EXPORTER_OTLP_HEADERS"] = os.getenv(
	"OTEL_EXPORTER_OTLP_HEADERS", f"x-mlflow-experiment-id={experiment_id}"
	)
	if "OTEL_EXPORTER_OTLP_HEADERS" not in os.environ:
	os.environ["OTEL_EXPORTER_OTLP_HEADERS"] = f"x-mlflow-experiment-id={experiment_id}"

---

Was this helpful? React with 👍 or 👎 to provide feedback.