Build once, deploy twice: HPE integrates Aruba and Juniper without merging them

Build once, deploy twice: HPE integrates Aruba and Juniper without merging them

HPE says it has done in one year what its competitors needed a decade for: merging two major networking portfolios, Aruba and Juniper, without forcing either customer base to switch platforms. At HPE Discover in Las Vegas, Sudheer Matta, SVP Products at HPE Networking, laid out how that works in practice, and why the platforms themselves will keep existing for the next 10 years.

This strategy addresses one of the most challenging aspects of technology mergers: maintaining customer trust while delivering accelerated value. HPE’s guiding principles ensure no customer gets left behind, regardless of which platform they chose before the acquisition.

Build once, deploy twice architecture

The foundation of HPE’s integration strategy is the fortunate compatibility between the two platforms. Both Aruba Central and Juniper Mist were built as cloud-native, AI-native systems with microservice architectures. This architectural similarity enables HPE to develop features once and deploy them across both platforms without forcing customers to change their operations.

Matta says: “We’re building 1 feature, and the feature is going to be deployed on Aruba or on Mist.”

This approach means Aruba customers can continue using Aruba Central, the CLI they know, and their existing management practices. Similarly, Juniper Mist customers maintain their familiar experience. Yet both customer bases benefit from the combined innovation of a much larger R&D organization.

Also read: HPE Networking firmly takes the wheel of the self-driving network

Feature parity timeline and customer commitment

HPE is moving methodically to bring feature parity between the Aruba and Mist platforms. The company has already migrated large experience models that measure Teams and Zoom performance from Mist to Aruba Central. Additional capabilities, including agentic interfaces and self-driving actions, are following a phased rollout.

“I would say slowly, I think in the next 6 to 12 to 18 months, more and more of the models will converge,” Matta explained. The timeline reflects HPE’s commitment to earning customer trust rather than rushing features to market that could fail in production environments.

Dual-platform hardware already shipping

HPE has moved beyond strategy to execution with the hardware part of the business. The company announced and is currently shipping dual-platform access points that work on both Aruba Central and Mist. Customers can deploy the same hardware on either management platform, and even switch between platforms without changing the physical infrastructure.

“We’ve announced new access points that today work on Aruba and Mist. You can switch. You could basically take an access point, you say, ‘Hey, I want to deploy a Mist network.’ Same access point today works on that. And then, oh no, I want to deploy on Aruba network. Same access point.”

Self-driving networks in production today

While many vendors discuss AI and automation as future capabilities, HPE is running autonomous networks in production at some of the world’s most demanding environments. The company’s self-driving network technology is already deployed at major retailers, healthcare systems, logistics companies, and other Fortune 500 enterprises.

Matta drew a compelling parallel to Waymo’s autonomous vehicles to illustrate the trust required for autonomous systems. Just as Waymo built trust through 200 million miles without causing a fatality, HPE has built network automation trust through years of production deployments.

Measurable customer outcomes

The results of self-driving networks are quantifiable and dramatic. The world’s largest retailer, which previously deployed 300-500 access points per night, now deploys 3,000 APs per night with HPE, a 10x improvement according to Matta. He also claimed that ServiceNow has seen a 90% reduction in user complaints after deploying HPE’s self-driving network.

“My assertion today with an HPE network, every single time, if I give a human 10 times to build a network and we give our AI 10 times to build the same network, 100% of the time we will build a better network than a human,” Matta stated. “Because humans cannot see packets in the air, we see packets in the air.”

The HPE Discover venue itself, the Venetian in Las Vegas, demonstrates this capability in action. The hotel operates the world’s largest single co-located 10,000 access point network, and it runs on HPE’s self-driving technology today.

The path goes from self-driving towards fully autonomous networking

While significant self-driving capabilities already exist in production, HPE continues working toward fully autonomous networking. The vision involves customers simply providing intent, such as “I have medical devices, guests, doctors, and IoT devices”, and then stepping away from the keyboard while the network configures and optimizes itself.

Market opportunity through cross-pollination

Beyond technical integration, Matta identified a significant market opportunity in cross-pollination between customer bases. Many HPE customers who purchase servers, storage, and hybrid IT solutions haven’t adopted HPE networking. Conversely, many networking customers haven’t explored HPE’s broader infrastructure portfolio.

“Being able to cross-pollinate between these two customer bases is the immense opportunity ahead of us”, Matta noted.

This future builds on the trust HPE has established through current deployments. Just as Waymo needed thousands of successful rides before mainstream adoption, autonomous networking requires proven reliability in demanding production environments. With major enterprises already running mission-critical operations on self-driving networks, HPE is building that foundation of trust. HPE Networking has already made significant steps forward in this merger, but there is more to come, especially around the fully autonomous networking, which can automate even more manual network administrator work.