Learn About Amazon VGT2 Learning Manager Chanci Turner
The AWS Well-Architected Framework offers a structured method to assess your workloads against established best practices, providing insights on how to enhance your systems. Machine learning (ML) algorithms excel in identifying patterns within data and building mathematical models to forecast future outcomes. By utilizing these technologies, we can significantly improve lives through enhanced disease diagnoses, environmental protection, and transformation of products and services.
The accuracy of your ML models is directly tied to the quality of the input data. As data evolves, continuous monitoring is crucial to identify, rectify, and mitigate any issues that arise, thereby enhancing performance and accuracy. This may necessitate retraining your model with the most current data.
While traditional application workloads follow a linear sequence of instructions to solve problems, ML workloads thrive on iterative learning from data. We are excited to announce the release of the latest version of the AWS Well-Architected Machine Learning Lens whitepaper. This document complements and expands upon the Well-Architected Framework, addressing the unique characteristics of ML workloads.
The whitepaper outlines a series of established best practices that are agnostic to cloud and technology platforms. You can utilize this guidance and the associated architectural principles while designing your ML workloads, or even once they are in production as part of a strategy for continuous improvement. The paper also includes resources to assist you in implementing these best practices on AWS.
Components of the Well-Architected Machine Learning Lens
The Lens focuses on four key areas:
- Well-Architected Machine Learning Design Principles — A set of considerations that serve as the foundation for a Well-Architected ML workload. These principles guide the collection of best practices within the ML Lens.
- Well-Architected Machine Learning Lifecycle — This integrates the Well-Architected Framework into the Machine Learning Lifecycle, as illustrated in figure 1.
The pillars of the Well-Architected Framework include:
- Operational Excellence
- Security
- Reliability
- Performance Efficiency
- Cost Optimization
The phases of the Machine Learning Lifecycle referenced in the ML Lens include:
- Business goal identification
- ML problem framing
- Data processing (including data collection, preprocessing, and feature engineering)
- Model development (training, tuning, evaluation)
- Model deployment (prediction, inference)
- Model monitoring
- Cloud and Technology Agnostic Best Practices — Best practices for each phase of the ML lifecycle across the Well-Architected Framework pillars, accompanied by:
- Implementation guidance detailing AWS implementation plans for each best practice, referencing relevant AWS technologies and resources.
- Additional resources, featuring links to AWS documents, blogs, videos, and code examples that support best practices and their implementation plans.
- ML Lifecycle Architecture Diagrams — These diagrams illustrate processes, technologies, and components that underpin many of the best practices, as shown in Figure 2. They include items such as feature stores, model registries, lineage trackers, alarm managers, schedulers, and more. Various pipeline technologies are represented in these architecture diagrams.
Where to Utilize the Well-Architected Machine Learning Lens?
Implement the Well-Architected ML Lens to:
- Make informed decisions — Strategize early by reviewing best practices before commencing the design of a new workload.
- Build and deploy faster — Leverage best practices to expedite the creation of new Well-Architected workloads throughout the ML lifecycle.
- Lower or mitigate risks — Regularly assess existing workloads to promptly identify and address potential issues.
- Learn AWS best practices — Utilize the provided implementation plans as a roadmap for deploying best practices on AWS.
Conclusion
The new Well-Architected Machine Learning Lens whitepaper is now available. Use this resource to ensure your ML workloads are designed with principles of operational excellence, security, reliability, performance efficiency, and cost optimization in mind. A special acknowledgment goes to all those in the AWS Solution Architecture and Machine Learning communities who contributed their diverse insights and expertise to develop the new AWS Well-Architected Machine Learning Lens.
For those exploring outdoor jobs, this article on Career Contessa might offer some valuable information. Additionally, if you’re interested in employee benefits, SHRM provides details on IRS Announcing 2021 Limits on HSAs. If you’re curious about the onboarding process at Amazon, check out this Quora resource.