Machine Learning Lens

Publication date: July 5th, 2023 (Document history)

In recent years, machine learning (ML) has moved from research and development to the mainstream, driven by the increasing number of data sources and scalable cloud-based compute resources. AWS’ customers currently use AI/ML for a wide variety of applications such as call center operations, personalized recommendations, identifying fraudulent activities, social media content moderation, audio and video content analysis, product design services, and identity verification. Industries using AI/ML include healthcare and life sciences, industrial and manufacturing, financial services, media and entertainment, and telecom.

Machine learning, through its use of algorithms to find patterns in data, can bring considerable power to its customers and thus recommends responsibility in its use. AWS is committed to developing fair and accurate AI and ML services and providing you with the tools and guidance needed to build AI and ML applications responsibly. For more information on this important topic, refer to AWS' Responsible AI.

This whitepaper provides you with a set of proven best practices. You can apply this guidance and architectural principles when designing your ML workloads, and after your workloads have entered production as part of continuous improvement. Although the guidance is cloud- and technology-agnostic, the paper also includes guidance and resources to help you implement these best practices on AWS.

Introduction

The AWS Well-Architected Framework helps you understand the benefits and risks of decisions you make while building workloads on AWS. By using the Framework, you learn operational and architectural best practices for designing and operating workloads in the cloud. It provides a way to consistently measure your operations and architectures against best practices and identify areas for improvement.

Your ML models depend on the quality of input data to generate accurate results. As data changes with time, monitoring is required to continually detect, correct, and mitigate issues with accuracy and performance. This monitoring step might require you to retrain your model over time using the latest refined data.

While application workloads rely on step-by-step instructions to solve a problem, ML workloads enable algorithms to learn from data through an iterative and continuous cycle. The ML Lens complements and builds upon the Well-Architected Framework to address this difference between these two types of workloads.

This paper is intended for those in a technology role, such as chief technology officers (CTOs), architects, developers, data scientists, and ML engineers. After reading this paper, you will understand the best practices and strategies to use when you design and operate ML workloads on AWS.