How HPE Zerto built an agentic troubleshooting system with Amazon Bedrock
Machine Learning Blog
This article describes how HPE Zerto built an agentic troubleshooting system powered by Amazon Bedrock to help disaster recovery and data protection teams resolve operational issues faster.
- Multi-agent architecture with Orchestrator, ZVM, and VRA sub-agents for specialized troubleshooting tasks
- Deployed on-premises as a pod in Zerto product with access to live environment data via Model Context Protocol servers
- Uses Amazon Bedrock for model inference, Guardrails for content security, and Knowledge Bases for retrieval-augmented generation
- Evaluated multiple foundation models using PyTest and Strands Agents Evals SDK with basic, multi-turn, and dynamic test suites
- 20% customer adoption since Q2 2026 release with 10% reduction in support cases for supported workflows
- Real-time streaming via Server-Sent Events improves user experience by showing investigation progress
The system transforms fragmented resilience data into actionable insights, helping teams reduce troubleshooting time and improve disaster recovery readiness without adding operational burden.
The AWS News Feed is currently looking for gold sponsors. If you want to support the AWS community and reach a large audience of AWS professionals, consider sponsoring the AWS News Feed.
Related articles
The AWS News Feed is currently looking for silver sponsors. If you want to support the AWS community and reach a large audience of AWS professionals, consider sponsoring the AWS News Feed.