ONUG Triple T presentations provide you with a unique opportunity to view Tools, Technologies and Techniques that map to solutions that enable trusted Infrastructure to drive business value in today’s Enterprise. From real-world NetDevOps to automating and securing multi-cloud deployments, you will find answers to your unique requirements and challenges.
In this session, Daniel Conroy and Jason Redman will explore how artificial intelligence is being used to improve human outcomes—starting with veteran mental health. They’ll discuss how AI-driven insights and automation can help identify at-risk individuals earlier, connect them to care faster, and strengthen collaboration between government, private, and nonprofit partners. The conversation will also highlight ethical and operational lessons learned from applying AI responsibly in high-stakes, human-centered environments.
The idea of self-healing, autonomous networks has always been a tantalizing goal. While businesses and technology leaders have been talking about it for a long time, true autonomous networks remain elusive in production. AI is reviving the idea of autonomous networks with new capabilities, but at the same time “AI washing” makes it difficult for businesses to tell what’s real. In this session we’ll assess the state of autonomous networking and look at the emerging technological elements which will turn it from science fiction into…
The workshop will walk through the Aviz Agent SDK (open source) and different implementation options (interactive agents, background agents). Agent Integration patterns with data sources/ connectors Best practices for agent design, testing, version control Deployment Options As part of the workshop, we will walk through 3 different agent examples and show the end-to-end workflow and agent use leveraging our Aviz AI ONE Center.
The workshop will provide insights into How to apply AI platform capabilities (data connectors/ MCP, LLMs, agents, contextual data/ RAG) to a set of very common NetOps use cases in a real multi-vendor network environment Explore different approaches on how to collaborate while bringing in a Network Copilot as team member Showcase examples from our Hackathon covering SONiC Network operations and troubleshooting, general network support flows for multi-vendor OS environments, and granular real-time observability of network traffic
AI agents are redefining what’s possible in network automation. In this session, we’ll show how the open-source Aviz Agent SDK makes it easy to extend your scripts with AI, whether you want to build your own agent or integrate existing ones. You’ll learn: The two main agent models: interactive copilots and background agents Simple integration patterns with your existing data sources and connectors Best practices for agent design, testing, and version control Deployment options that fit different NetOps environments We’ll close with a showcase of…
In just 120 minutes, IT executives get a strategic perspective and competitive edge to design and refine their AI adoption roadmap using a practical framework that accelerates ROI and future-proofs your AI strategy. The workshop will walk through: Aligning Agentic AI with Business Vision, Mission, and Growth Objectives Value Engineering: Core Principles for Developing Brand-Centric AI Cultural Preservation and Ethical AI Governance Human-AI Collaboration Frameworks Value Engineering for High-Performance Workflows Participants will walk away equipped with a toolkit to scope their agentic project portfolio and…
Picture this: coffee in hand, you ask, “Good morning, NORM. Where should I focus today?” Instantly, NORM replies, “Good morning. Seattle’s router is dropping packets above 10%. Help desk tickets are piling up there—start there first.” In this session, meet NORM – your Network Observability Remediation Manager. NORM represents the “art of the possible” when you power it with intelligent network observability data and analytics, NORM becomes your AI-driven wingman that can triage issues, open or escalate tickets, roll back configurations, and even predict utilization…
AI workloads are only as fast as the networks that carry them. Yet enterprises often discover congestion, incast, and collective operation bottlenecks only after GPUs are deployed—when fixes are costly. Keysight’s KAI Data Center Builder provides a smarter path: it emulates AI workloads and benchmarks backend fabrics without consuming GPU cycles. This enables IT teams to compare architectures, validate performance, and de-risk investments before scaling. Attendees will leave with a repeatable validation framework they can apply immediately to ensure their AI infrastructure is ready to…
