Amazon Bedrock Agents → OpenObserve

Automatically capture latency and invocation metadata for every Amazon Bedrock Agent call using the OpenInference Bedrock instrumentation.

Prerequisites

Python 3.8+
An OpenObserve account (cloud or self-hosted)
Your OpenObserve organisation ID and Base64-encoded auth token
AWS credentials with AmazonBedrockFullAccess permissions
A Bedrock Agent with an Agent ID and Alias ID. See Create a Bedrock agent.

Installation

pip install openobserve-telemetry-sdk openinference-instrumentation-bedrock boto3 python-dotenv

Configuration

Create a .env file in your project root:

OPENOBSERVE_URL=https://api.openobserve.ai/
OPENOBSERVE_ORG=your_org_id
OPENOBSERVE_AUTH_TOKEN=Basic <your_base64_token>

AWS_ACCESS_KEY_ID=your-access-key-id
AWS_SECRET_ACCESS_KEY=your-secret-access-key
AWS_DEFAULT_REGION=your-aws-region
BEDROCK_AGENT_ID=your-agent-id
BEDROCK_AGENT_ALIAS_ID=TSTALIASID

Instrumentation

Call BedrockInstrumentor().instrument() before creating any boto3 client.

from dotenv import load_dotenv
load_dotenv()

from openinference.instrumentation.bedrock import BedrockInstrumentor
from openobserve import openobserve_init

BedrockInstrumentor().instrument()
openobserve_init(resource_attributes={"service.name": "amazon-bedrock-agents"})

import os, uuid, boto3

client = boto3.client(
    "bedrock-agent-runtime",
    region_name=os.environ.get("AWS_DEFAULT_REGION", "us-east-1"),
)

response = client.invoke_agent(
    agentId=os.environ["BEDROCK_AGENT_ID"],
    agentAliasId=os.environ.get("BEDROCK_AGENT_ALIAS_ID", "TSTALIASID"),
    sessionId=str(uuid.uuid4()),
    inputText="What can you help me with?",
)
completion = ""
for event in response.get("completion", []):
    if "chunk" in event:
        completion += event["chunk"]["bytes"].decode("utf-8")
print(completion)

What Gets Captured

Attribute	Description
`operation_name`	Always `bedrock_agent.invoke_agent`
`openinference_span_kind`	Always `AGENT`
`llm_provider`	Always `aws`
`input_value`	Text sent to the agent
`output_value`	Agent response text
`output_mime_type`	Always `text/plain`
`span_status`	`OK` on success, `ERROR` on failure
`duration`	End-to-end invocation latency

Viewing Traces

Log in to OpenObserve and navigate to Traces in the left sidebar
Click any bedrock_agent.invoke_agent span to inspect latency and input/output content

Amazon Bedrock Agents trace in OpenObserve

Next Steps

Track agent response times, monitor failure rates, and correlate agent spans with the rest of your application traces.