# Aviz Networks panel examines AI observability and SONiC

URL: https://decisioninsights.ai/aviz-networks-panel-examines-ai-observability-and-sonic/

Leaders from Carahsoft, AvidThink, and Aviz Networks discussed how AI-driven observability, open networking with SONiC, and vendor-agnostic automation are changing network operations. The topic matters to enterprise IT leaders as bandwidth demands and AI adoption increase.Operational ContextPanelist Ram Mohan Hariprasad said networking must scale quickly to support GPU and DPU acceleration, rising bandwidth requirements, and generative AI adoption. He noted that compute hardware advances arrive every few months, so networking must adapt at the hardware level as well as in operations and observability.AI-Driven ObservabilityThe panel said traditional observability stacks built on dashboard sprawl and manual triage are no longer sufficient. Aviz described a move beyond click-based operations toward large language models that support real-time, conversational queries. Users can ask questions and receive contextualized answers with charts and graphs, which the blog says reduces time-to-insight across multi-vendor, multi-domain environments.Flexibility and Cost ConsiderationsPanelists identified a shared priority: enterprises want flexibility without budget increases. Roy Chua said, “It's not just about speed; it's about flexibility and intelligence at the infrastructure layer.” The panel described a standardized operational layer powered by Aviz ONES as a way to pursue that goal.SONiC and Open NetworkingScott Robohn described open networking as an extension of the Linux model, saying, “What Linux did for servers, SONiC is doing for switches.” Ram Mohan Hariprasad said SONiC now functions as a standardized foundation comparable to Linux in compute. Aviz extends SONiC with observability, automation, and enterprise support to connect community development with production readiness.Data Layer StandardizationThe panel identified the next phase of standardization at the data layer, where normalizing telemetry, logs, and metrics across hardware and operating systems would allow AI to manage networks more effectively. Aviz describes this approach as a single AI for diverse environments that provides centralized intelligence. Roy Chua said SONiC's success rests on hyperscaler deployments running at scale, which gives enterprises confidence.The panel's overall message is that AI-ready networks will depend on open platforms, standardized telemetry, and unified observability. This Blog Summary is a fact-based summary of the vendor blog, provided for enterprise IT, security, and networking decision-makers evaluating how these approaches affect infrastructure operations.
