HPE Self-Driving Network Autonomously Fixes Up to 80% of Issues
Hewlett Packard Enterprise has deployed a self-driving network capability that detects and resolves 70% to 80% of potential network issues without human intervention, Fierce Network reported. Rami Rahim, executive vice president, president and general manager of networking at HPE, stated during the company’s Networking Investor Day that the system informs IT staff of fixes only after completing them. The technology addresses common disruptions such as poor performance, slow connections, roaming failures, and DNS problems.
Marvis AI Engine Deployment and Customer Trust Milestones
The system relies on the Marvis AI engine to screen every customer service ticket, automatically resolving known issues or storing data for future automated repairs. According to Sujai Hajela, executive vice president and general manager of HPE’s campus and branch business, customers authorized approximately 10,000 self-driving actions across HPE cloud platforms within a 90-day window. HPE deploys the capability incrementally to build confidence, with users activating features one by one rather than ceding complete network control immediately.
Active Threat Response and Regulatory No-Fly Zones
Despite high levels of automation, human operators remain essential for managing security incidents involving active attackers and executing configuration changes tied to regulatory compliance. Hajela explained that HPE established designated no-fly zones where automation cannot operate, requiring explicit customer sign-off and rollback capabilities for any changes impacting regulated industries like healthcare. Rival equipment maker Cisco is pursuing a parallel strategy, having expanded its AgenticOps model in February to include autonomous troubleshooting for campus, branch, and industrial networks.
Surging Bot Traffic and Telecom Carrier AI Strategies
Rising volumes of automated web activity increase the operational urgency for autonomous infrastructure management. Cloudflare data from May indicates that non-human bot traffic surpassed human traffic, reaching nearly two-thirds of all internet traffic. Rahim noted that because non-human traffic cannot report connection failures, networks must rely on continuous user and device experience measurements rather than basic status monitoring. Meanwhile, telecommunications carriers are pursuing monetization strategies that include deploying GPU-as-a-service offerings outside the United States to capture sovereign AI demand.
Frequently Asked Questions
How does HPE calculate the 70-80% autonomous fix rate?
HPE tracks the metric by logging every instance where the Marvis AI engine intercepts a customer service ticket and successfully diagnoses and repairs the underlying network issue without human intervention.
What specific hardware agreements drove HPE’s updated networking forecast?
HPE secured a $1.2 billion deal with Vultr on September 30 for AMD Helios hardware, which Rahim projected will generate more than $1 billion in networking opportunities over two years, alongside an Oracle deployment agreement announced September 2.
Who acquired HPE’s Telco Solutions business unit?
HCLTech completed its acquisition of HPE’s Telco Solutions unit on August 5, eight months after the initial agreement, shifting HPE’s carrier focus toward products such as PTX12000 routers and agentic Routing Director features.
Worth a look