Network Observability and Interop in the Age of AI: Insights from Industry Experts

AI is reshaping both how networks operate and what they’re expected to support. In a recent expert panel, leaders from Carahsoft, AvidThink, and Aviz Networks discussed how AI-powered observability, open networking with SONiC, and vendor-agnostic automation are enabling smarter, faster, and more cost-efficient infrastructure.

Ram Mohan Hariprasad opened with a sweeping view of today’s rapid evolution. From GPU and DPU acceleration to surging bandwidth demands and generative AI adoption, networking must scale—fast.

Every few months, we see breakthroughs in compute hardware. Networking has to adapt—not just at the hardware level, but across operations and observability too.

Roy Chua emphasized the shift isn’t just technical—it’s deeply operational:

It’s not just about speed; it’s about flexibility and intelligence at the infrastructure layer.

Traditional observability stacks, built around dashboard sprawl and manual triage, are no longer enough. Ram described how Aviz is moving beyond the “click-ops” model using large language models (LLMs) for real-time, conversational observability.

You don’t have to hunt across ten dashboards anymore. Instead, you ask questions, and AI agents deliver contextualized answers—complete with pie charts, time series graphs, or honeycomb overlays.

This AI-driven model reduces time-to-insight and creates a single pane of intelligence across multi-vendor, multi-domain environments.

One theme emerged clearly: enterprises want flexibility—without budget explosions.

Enterprises want to embrace change—new hardware, new speeds—but without rewriting apps or overspending.

With a standardized operational layer (powered by Aviz ONES), organizations can:

Scott Robohn framed the shift to open networking as a continuation of the Linux story:

What Linux did for servers, SONiC is doing for switches.

Ram expanded on this, explaining how SONiC now serves as a standardized foundation, much like Linux in the compute world.

Aviz extends SONiC’s value with observability, automation, and enterprise-grade support—bridging the gap between community innovation and production readiness.

Roy Chua pointed out:

SONiC's success is grounded in hyperscaler proof points—it’s running at scale today. That gives enterprises real confidence.

The panel closed with a look toward the future—and what AI-native infrastructure really means.

The next phase of standardization is at the data layer. Normalizing telemetry, logs, and metrics across hardware and OS types is what will allow AI to actually manage the network effectively.

With this foundation, Aviz enables a “single AI for diverse environments”—centralized intelligence without operational bloat.

Final Words

As Scott aptly summarized:

The future isn’t just about speed—it’s about smart infrastructure that adapts, interoperates, and learns.

Backed by hyperscaler DNA, community-driven innovation, and an enterprise-first mindset, Aviz and its partners are building the next generation of open, intelligent, and AI-ready networks.