AWS wellarchitected documentation change
Summary
Added review guidance for comparing managed vs self-managed generative AI services, including compliance benchmarking for regulated workloads
Security assessment
Focuses on cost-performance optimization and compliance requirements without addressing specific security vulnerabilities or introducing security controls. Mentions governance but lacks concrete security mechanisms or threat mitigation.
Diff
diff --git a/wellarchitected/latest/financial-services-industry-lens/fsicost17.md b/wellarchitected/latest/financial-services-industry-lens/fsicost17.md index d2c988fdb..8891684da 100644 --- a//wellarchitected/latest/financial-services-industry-lens/fsicost17.md +++ b//wellarchitected/latest/financial-services-industry-lens/fsicost17.md @@ -14,0 +15,6 @@ There are multiple factors that affect the architecture, for example, new enhanc +**Review** + +Continuously re-assess whether managed generative AI services (for example, Amazon Bedrock or Amazon Q) or self-managed open-model stacks offer the best price-performance and governance balance for your risk and compliance constraints. + +For highly regulated workloads, periodically benchmark in-house fine-tuned models against Bedrock-hosted foundation models to verify that the chosen deployment pattern continues to meet cost, latency, and compliance requirements. +