AgentsSeptember 8, 2026via AWS Machine Learning Blog
How HPE Zerto built an agentic troubleshooting system with Amazon Bedrock
Why it matters
A real-world agent deployment case study showing the engineering patterns practitioners need to move agents from experiment to production: multi-agent orchestration, on-premises grounding, and integration with live operational data.
Key signals
- Multi-agent architecture (specific agents not named in headline, but implied in 'multi-agent')
- On-premises deployment model
- Built with Strands Agents framework
- Amazon Bedrock backend
- Grounding agents in live disaster recovery data
- Production troubleshooting system (not pilot)
- Multi-agent architecture for disaster-recovery troubleshooting
- On-premises deployment model (agents run inside customer environment)
- Powered by Amazon Bedrock
- Grounding agents in live DR data (engineering challenge described)
- HPE Zerto is the platform (disaster-recovery/backup software vendor)
The hook
HPE Zerto deployed multi-agent troubleshooting in production — not a pilot. Here's how they grounded agents in live DR data on-premises.
HPE Zerto built an agentic troubleshooting system powered by Amazon Bedrock that runs on-premises inside the customer environment. This post describes the multi-agent architecture, the on-premises deployment model built with Strands Agents, and the engineering challenges of grounding agents in live …