Amazon SageMaker

Overview

This guide shows you how to integrate Latitude Telemetry into an application that invokes models hosted on Amazon SageMaker through boto3.

You’ll keep calling SageMaker exactly as you do today. Telemetry simply observes and enriches those calls.

Auto-instrumentation for SageMaker is available in the Python SDK. For other languages, send traces via the OpenTelemetry exporter.

Requirements

A Latitude account and API key
A Latitude project slug
A project that uses boto3 with a deployed SageMaker endpoint
AWS credentials (AWS_ACCESS_KEY_ID, AWS_SECRET_ACCESS_KEY, AWS_REGION) and a SAGEMAKER_ENDPOINT_NAME

Steps

Install

pip install latitude-telemetry boto3

uv add latitude-telemetry boto3

poetry add latitude-telemetry boto3

Initialize and use

import json
import os

import boto3

from latitude_telemetry import Latitude, capture

latitude = Latitude(
    api_key="your-api-key",
    project="your-project-slug",
    instrumentations={"sagemaker": boto3},
)

client = boto3.client("sagemaker-runtime", region_name="eu-central-1")

def generate_reply():
    response = client.invoke_endpoint(
        EndpointName=os.environ["SAGEMAKER_ENDPOINT_NAME"],
        ContentType="application/json",
        Body=json.dumps({"inputs": "Hello", "parameters": {"max_new_tokens": 50}}),
    )
    return json.loads(response["Body"].read().decode())

capture("generate-reply", generate_reply)

latitude.shutdown()

Seeing Your Traces

Once connected, traces appear automatically in Latitude:

Open your project in the Latitude dashboard
Each execution shows input/output messages, model, token usage, latency, and errors

Replicate IBM watsonx.ai

⌘I

Overview

Getting Started

Observe

Understand

Refine

Security and Compliance

Deployment

Development

More

Overview

Requirements

Steps

Seeing Your Traces

​Overview

​Requirements

​Steps

​Seeing Your Traces

Overview

Requirements

Steps

Seeing Your Traces