Skip to main content

​
Overview

traceAI-pipecat adds OpenTelemetry observability to Pipecat applications, enabling tracing and monitoring of voice agents with automatic attribute mapping to Future AGI conventions. It works by mapping Pipecat’s OpenTelemetry spans (LLM, STT, TTS, tools, sessions) into Future AGI’s tracing and evaluation format so you can observe and evaluate your pipeline in the Future AGI platform.

Source Repository

Source code, examples, and issues for the Pipecat integration

PyPI Package

The traceAI-pipecat package on PyPI

Future AGI

Learn more about the Future AGI observability and evaluation platform

​
Installation

This is a community-maintained package distributed separately from pipecat-ai: 
pip install traceAI-pipecat
For gRPC transport, also install the gRPC extras: 
pip install "fi-instrumentation[grpc]"

​
Prerequisites

​
Future AGI Account Setup

Before using the integration, you need:
  1. Future AGI Account: Sign up at Future AGI
  2. API Credentials: An API key and secret key from your Future AGI account

​
Required Environment Variables

  • FI_API_KEY: Your Future AGI API key
  • FI_SECRET_KEY: Your Future AGI secret key
import os

os.environ["FI_API_KEY"] = FI_API_KEY
os.environ["FI_SECRET_KEY"] = FI_SECRET_KEY

​
Setup

Enable observability by registering a tracer provider and then installing the Pipecat attribute mapping. The mapping updates your existing span exporters to convert Pipecat attributes into Future AGI conventions. 
from fi_instrumentation.otel import register, Transport, ProjectType
from traceai_pipecat import enable_http_attribute_mapping

# Register the tracer provider to establish the observability pipeline
trace_provider = register(
    project_type=ProjectType.OBSERVE,
    project_name="My Pipecat Voice App",
    transport=Transport.HTTP,
    set_global_tracer_provider=True,
)

# Install attribute mapping to convert Pipecat attributes to Future AGI conventions
enable_http_attribute_mapping()

# Your Pipecat application code here...
For gRPC transport, use enable_grpc_attribute_mapping() instead, or specify the transport explicitly: 
from traceai_pipecat import enable_fi_attribute_mapping
from fi_instrumentation.otel import Transport

enable_fi_attribute_mapping(transport=Transport.HTTP)  # or Transport.GRPC
Call register() before installing attribute mapping, and make sure the transport type matches your tracer provider configuration. See the source repository for the full API reference, feature list, and troubleshooting.

​
Compatibility

The integration targets Pipecat’s OpenTelemetry tracing support. A specific tested Pipecat version is not stated in the repository; check the source repository for the latest compatibility details and changelog.