Operations | Monitoring | ITSM | DevOps | Cloud

Everett Sinclair, Network Administrator at Conway Corporation, explains why Kentik became their “one pane of glass” for cloud-based network visibility, rapid troubleshooting, and smarter peering and caching decisions. With Kentik’s SaaS network intelligence platform, Conway gets updates automatically, avoids server rebuilds, and can deploy cloud agents remotely to run simple metric tests close to customer locations.

AI is changing network operations (NetOps) from static automation into adaptive, data-driven systems that can summarize incidents, retrieve knowledge, and guide remediation with human oversight. In this talk, Phil Gervasi breaks down what “AI for NetOps” really means in practice, including the difference between classical ML and large language models (LLMs), why data pipelines matter more than model tuning, and how patterns like RAG (retrieval augmented generation), text-to-SQL, and agentic workflows turn raw telemetry into decisions.

Five customers explain why Kentik is essential for understanding traffic, controlling network cost, and planning for growth. Hear from Sorin Esanu (Race Communications), Michael Leclaire (MetroNet), John Lubeck (Midco), Wallace Lee (Imperva), and Everett Sinclair (Conway Corporation) on how they use Kentik to see “the bits on the wire,” dig deeper than traditional reporting tools, and turn network data into better customer experiences.

Three service provider customers share how Kentik AI Advisor helps them move faster, troubleshoot smarter, and put network data in more hands across the team. Hear from Everett Sinclair (Conway Corporation), Michael Leclaire (MetroNet), and John Lubeck (Midco) on why they chose Kentik to unify flow analytics, baselining, and anomaly detection in one platform, and how Kentik AI features make it easier to explore, explain, and act on what’s happening in their networks.

"When we started, maybe ten percent of our traffic went to a CDN. Now it's ninety percent on the CDNs or local caches and ten percent goes to transit. We've really flipped the whole script." — John Lubeck, Midco.

A BGP session goes down because a transit provider exceeded the maximum prefix limit. How do you find the root cause — fast? In this 10-minute demo, we walk through two approaches using Kentik AI Advisor. First, we troubleshoot step by step using natural language: asking AI Advisor to identify the affected interface, check for interface flapping, and review syslog messages until we find the maximum-prefix violation. Then we show how custom network context and natural language runbooks let AI Advisor do the entire investigation autonomously — following the same four steps a senior engineer would.