VIAVI at Tech Field Day 2025: Part 1 - Assessing the Current State of AI-driven Packet Analysis with VIAVI

As network environments grow in complexity, speeds, and feeds, packet-based troubleshooting gets increasingly complex. In this session, we explore how artificial intelligence has the potential to change the game—automating anomaly detection, accelerating root cause analysis, and uncovering patterns that might otherwise go unnoticed. Ward Cobleigh and Chris Greer examine how today’s Large Language Models (LLMs) handle PCAP data, sharing results from a controlled test using a small file with a clear anomaly. Models including ChatGPT, Claude Sonnet, Microsoft Copilot, and Google Gemini were evaluated for their ability to identify a 132-second server response delay. The results reveal a wide range of capabilities—from vague or hallucinated answers to accurate and actionable analysis. The session also explores real-world uses of AI for filtering, context enrichment, and filter creation, while explaining current limitations like data size restrictions and the need for secure data handling. Whether you're evaluating AI for daily triage or deeper forensics, this session delivers practical insights into what’s possible now—and what’s coming next. Watch now to see how today’s popular AI tools handle packet analysis—and what you need to know before trusting them as part of your workflow.