利用人工智能提高人道主义工作负荷恢复能力.pdf

1、 2025,Amazon Web Services,Inc.or its affiliates.All rights reserved.2025,Amazon Web Services,Inc.or its affiliates.All rights reserved.A I M 3 3 6Using AI to improve humanitarian workload resilienceMike George(he/him)Principal Solutions ArchitectAmazon Web ServicesMatt Bollinger(he/him)Senior Soluti

2、ons ArchitectAmazon Web Services 2025,Amazon Web Services,Inc.or its affiliates.All rights reserved.2025,Amazon Web Services,Inc.or its affiliates.All rights reserved.A B I L I T Y O F A W O R K L O A D T O R E C O V E R F R O M I N F R A S T R U C T U R E O R S E R V I C E D I S R U P T I O N SResi

3、lienceThe mental modelContinuous improvement CI/CD,observability,moving beyond pre-deployment testing towards chaos engineering patternsHigh availabilityResistance to common failures through design and operational mechanisms at a primary site Core services,design goals to meet availability goalsDisa

4、ster recoveryReturning to normal operation within specific targets at a recovery site for failures that cannot be handled by HABackup and recovery,data bunkering,managed recovery objectives 2025,Amazon Web Services,Inc.or its affiliates.All rights reserved.Categories of failure(SEEMS)Single points o

5、f failureLack of redundancy or fault toleranceExcessive loadNot havingsufficient capacity/resources/limitsExcessive latencyNot responding in the expected timeMisconfiguration and bugsIncorrect executionShared fateViolating intended fault isolation 2025,Amazon Web Services,Inc.or its affiliates.All r

6、ights reserved.Agentic AIAgentUsersFoundation ModelToolsMCP 2025,Amazon Web Services,Inc.or its affiliates.All rights reserved.Agentic resilienceAWS CloudUserAmazon BedrockAgentToolsuse_awscalculate_letter_gradeAWS DocumentationCloud Control APIAWS Documentati