Introduction to SageMaker

Duration: 45 min

AWS SageMaker is a fully managed machine learning service that enables data scientists and developers to build, train, and deploy ML models quickly. This module covers the core components, Studio environment, notebook instances, and pricing model.

What is SageMaker?

SageMaker removes the heavy lifting from each step of the ML workflow. It provides integrated tools for data labeling, feature engineering, model training, hyperparameter optimization, and deployment—all in one platform.

Core Components

SageMaker Studio is the web-based IDE for ML development. Notebook Instances provide Jupyter environments for experimentation. Processing Jobs handle data preparation at scale. Training Jobs train models on distributed infrastructure. Endpoints serve real-time predictions. Pipelines orchestrate end-to-end ML workflows.

Architecture Overview

import sagemaker
from sagemaker import get_execution_role

# Initialize SageMaker session
session = sagemaker.Session()
role = get_execution_role()
bucket = session.default_bucket()

print(f"Default bucket: {bucket}")
print(f"Execution role: {role}")
print(f"Region: {session.boto_region_name}")

Pricing Model

SageMaker charges for compute resources used. Notebook instances are billed per hour. Training jobs charge for compute instances during training. Endpoints charge for hosting time. Processing jobs charge for compute resources. Feature Store charges for storage and ingestion.

# Check SageMaker pricing via AWS CLI
aws pricing get-products \
  --service-code AmazonSageMaker \
  --region us-east-1 \
  --query 'PriceList[0]' | jq .

Getting Started with SageMaker

# Create a simple notebook instance
from sagemaker.estimator import Estimator

# Define training parameters
training_params = {
    "image_uri": "382416733822.dkr.ecr.us-east-1.amazonaws.com/xgboost:latest",
    "role": role,
    "instance_count": 1,
    "instance_type": "ml.m5.xlarge",
    "output_path": f"s3://{bucket}/output"
}

print("SageMaker is ready for ML workflows")

Key Services Integration

SageMaker integrates with S3 for data storage, IAM for access control, CloudWatch for monitoring, and ECR for custom containers.

{
  "sagemaker_config": {
    "default_bucket": "my-sagemaker-bucket",
    "default_role": "arn:aws:iam::123456789012:role/SageMakerRole",
    "region": "us-east-1",
    "instance_types": ["ml.m5.xlarge", "ml.p3.2xlarge"],
    "max_runtime_seconds": 86400
  }
}

Quiz 1

Quiz 2

Quiz 3

Quiz 4

Quiz 5