Announcing AWS Well-Architected Agent, an AI-powered intelligence to optimize your cloud environment (preview)
~2 min read
devopsawswell-architectedaicost-optimizationinfrastructure-as-code
The Well-Architected console now includes an AI-powered agent that continuously monitors your environment and produces prioritized, goal-ranked recommendations with automation-ready remediation — available in preview.
In the Well-Architected console, choose Get started with Well-Architected Agent. You define an agent profile: which AWS accounts and applications to monitor, which optimization pillars matter (cost, performance, resilience, security), and the permissions required. You can scope by account, Region, and per-pillar goals.
Access control is clean: you provision customer-managed IAM roles the agent uses to read resource configurations, utilization metrics, and application topology. After you choose Get Started, resource and application recommendations are generated within 24 hours of profile creation.
For pre-deployment workloads, you can conduct an architecture review by uploading an IaC project — Terraform, AWS CloudFormation, or AWS CDK. Attach a .zip of the project, pick a Well-Architected lens, and the agent reviews the infrastructure before it ships. You can also define applications (accounts, Regions, services, tags, details) to add context that enhances recommendation relevancy.
The generated recommendations are prioritized across your resources and applications, ranked against your declared goals, with automation-ready remediation included. Each recommendation shows why the agent is suggesting it, impacts and trade-offs, and fixes across affected AWS resources.
Remediation comes in three forms: console actions, updated IaC templates with the exact code changes to copy into your codebase (e.g. a modified CDK function), or CLI commands — with step-by-step instructions you can roll out and verify. API access enables integration into your own tooling.
The Well-Architected Agent converts the framework from a periodic self-review exercise into continuous, goal-driven monitoring with ready-to-apply fixes. The IaC-template remediation is the standout feature — it turns "you should change this" into "here's the exact diff," which is the difference between a report and an outcome.