ONUG names Best in Show winners at AI Networking Summit in New York

ONUG recognized Alkira, Cisco, DevAI and BlueCat as Best in Show winners from its AI Networking Summit held Oct. 28–29 in New York City. The awards highlighted enterprise tools for AI infrastructure, networking, automation and security as ONUG prepares a live threat-testing event for its 2027 summit in Texas.

Why it matters: - ONUG’s awards spotlight the technologies enterprise IT leaders are treating as practical building blocks for AI at scale. - The winning solutions center on programmable networking, automation and security for AI-driven operations. - The summit also signaled a shift from AI demos to real-world validation and resilience testing.

What happened: - ONUG named four Best in Show winners from its AI Networking Summit, held Oct. 28–29 in New York City. - Nearly 1,750 attendees took part in the two-day event. - The winners were Alkira, a Lumen company, in AI Infrastructure; Cisco in Networking; DevAI in Automation; and BlueCat in Security. - The summit brought together enterprise IT practitioners, executives and technology innovators.

The details: - ONUG says the summit content was developed by its board, including leaders from HSBC, FedEx, Fidelity, RTX, Johnson & Johnson and Evernorth. - The event ran across five tracks: Infrastructure, Automation, Networking, Security and Post-Quantum Cryptography. - Sessions covered the AI-native enterprise network fabric, NOC/SOC convergence, zero trust and agentic operations. - Before the summit, ONUG Collaborative Working Groups challenged vendors to offer real options for an agentic control plane, autonomous infrastructure and an AI-enabled security operations center. - Fifteen companies accepted the challenge and submitted responses. - Best in Show went to the four solutions that best met the challenge. - Alkira won for Network Infrastructure on Demand and its programmable approach to enterprise networking. - Alkira and Lumen combine connectivity, routing, segmentation and visibility in one operating model. - Cisco won for its “Cisco AI Networking” session, led by Richard Licon, Director of Technical Marketing for Data Center and AI Networking. - Cisco showed how Nexus One and Cisco Cloud Control can unify the back-end GPU, front-end, storage and management fabrics of an AI cluster in one console across NX-OS, ACI and SONiC. - Cisco also highlighted AI Canvas for agentic cross-domain troubleshooting, 800G Smart Switches for distributed stateful enforcement and Isovalent eBPF integration for visibility from Kubernetes pods into the physical fabric. - DevAI won Best in Show for Automation for the second year in a row. - Dr. Susie Wee, DevAI co-founder and CEO, argued in her session that automation and AI depend on trustworthy, verified data underneath them. - BlueCat won for Security after presenting a scenario in which an AI agent runs six unscripted queries and leaves multiple network footprints without a single identity tying them together. - BlueCat showed how enterprises can link those traces into a single unit of work for security and operations teams.

Between the lines: - ONUG is framing AI networking as an operational problem, not just a technology showcase. - The vendor challenge suggests buyers want proof that agentic systems can be governed, observed and defended in production-like conditions. - Cisco’s focus on multiple fabrics and observability reflects the complexity of scaling AI infrastructure without losing control. - DevAI’s message and BlueCat’s demo both point to the same constraint: automation and agentic traffic are only useful if enterprises can trust the data and trace the activity.

What's next: - ONUG’s next AI Networking Summit is set for May 12–13, 2027, in Frisco, Texas, near Dallas. - ONUG plans to run a live Agentic 2.0 infrastructure at the 2027 summit and test it with exploits in real time. - Enterprise IT leaders, practitioners and solution providers can learn more at onug.net.

The bottom line: - ONUG is pushing the AI networking conversation from promise to proof, with awards now tied to how well vendors help enterprises operate AI safely at scale.