What’s New at Kentik: Platform Updates for April 2026
Summary
Over the past few months, we’ve been making the Kentik platform easier to use and more actionable, with AI increasingly at the center of how teams interact with it. AI Advisor sits near the middle of a lot of that progress, but this is not only an AI story. It’s also about Kentik becoming a more complete platform, bringing flow and NMS together in more useful ways, helping teams reduce network costs with sharper visibility, and making it easier to find the right workflows and get from question to answer faster.
AI Advisor is changing how networkers work
AI Advisor is one of the clearest examples of this shift in practice. AI Advisor is Kentik’s AI agent, built to work more like a virtual network engineer partner, helping teams reason through problems, pull the right data, interpret what it finds, and explain what’s going on without forcing users to stitch the whole investigation together by hand.
We’re already seeing customers use it in some very practical ways. Some are using it to accelerate incident triage when alerts fire, helping them get oriented faster and move more quickly toward root cause. Others are using it to answer hybrid and multi-cloud questions that are hard to untangle across teams and tools, especially when performance and cost are both in play. And some of the most common uses are the everyday questions that still eat time, the weird one-offs, the unfamiliar hosts, interfaces, circuits, or traffic patterns where you just want to get oriented and keep moving.
We’ve also recently launched a number of new capabilities around AI Advisor, including On-demand Connectivity Test, which lets AI Advisor run ping and traceroute directly from Universal Agents during an investigation, and the AI Advisor REST API, which brings the same investigation model into automation, chat, and other operational workflows. We’ve also continued expanding how teams can use AI Advisor through things like runbook enhancements, MCP server support, Slackbot access, and CSV export for tables, all aimed at making it easier to bring AI Advisor into the places where real work is already happening.
There’s enough there that it really deserves its own post, and we’ll dig into that more fully soon.
Kentik NMS capabilities have grown tremendously
One of the bigger platform stories for us has been how much Kentik NMS (Network Monitoring System) has evolved.
If you looked at Kentik’s NMS in the past and felt it still had some important gaps, it’s time to take a second look. Over time, we’ve been turning NMS into a much more complete operational system, with an all-new stateful alerting engine, broader telemetry support across syslog, traps, and streaming telemetry, and more workflow depth around how teams actually investigate and operate the network.
More importantly, these capabilities are not developing in isolation. We’re taking an AI-first, integrated approach to traditional network monitoring, combining NMS with Kentik’s flow visibility, synthetics, and AI-driven investigation to make it part of a more complete network intelligence platform rather than just another monitoring tool.
Syslog is a good example of how we’re connecting flow and NMS workflows in more useful ways. Syslog data is now available as Events telemetry in Data Explorer, allowing teams to analyze it with richer dimensions and views rather than treating it as something off to the side. We’ve also added a syslog-detected device filter in inventory, along with a Syslogs tab on device details, so if you’re already looking at a device and see something unusual, you can move straight into the relevant syslog context without bouncing into another tool or workflow.
What makes this especially valuable in Kentik is that syslog is not living in isolation. Because it can sit alongside flow and other telemetry in Data Explorer, teams can investigate operational issues with more context and less tool-hopping than they would get from a traditional NMS workflow. It is not just that Kentik now supports more NMS data. It’s that those signals are becoming part of a broader network intelligence workflow, where event data, traffic behavior, and device context can be analyzed together.
We’ve invested heavily in making our NMS one of the most modern and capable network monitoring solutions available, and we’re really just getting started.
Device Support Explorer removes guesswork before rollout
We’ve also been investing in device support for NMS and even added a new NMS Device Support Explorer, which gives users a much clearer view of what Kentik NMS collects for specific vendors and models without needing to try to onboard a device to find out.
We love it for the same reason customers do – it removes guesswork and trial-and-error. If you’re planning an expansion, verifying support across inventory, or deciding whether a new model provides the measurements you care about, that visibility is useful well before the first device is added.
Traffic Costs helps network designers really know what drives spending
Traffic Costs is another workflow we’ve been building out to help network teams get much clearer answers about what traffic is actually costing the business and what’s driving it. It gives operators, planners, and architects an automated way to calculate how different traffic slices contribute to connectivity costs, making it easier to turn traffic data into actionable insights for planning, optimization, and cost accountability.
Since its launch in 2025, we’ve continued expanding that workflow with more ways to break down spend. In addition to the original views by Source/Destination ASN, AS Group, AS Path, and Customer Port, teams can now analyze costs by CDN provider, OTT service and category, geography, and IP/CIDR. That gives teams a much more granular way to understand where costs are coming from and what to do about them, whether that means improving traffic engineering, tightening attribution, or identifying specific providers, services, or regions driving spend.
Finding the right thing faster
Another quiet, but genuinely useful improvement we’ve made lately is AI-powered Navigation Search.
As Kentik has grown, so has the number of workflows, screens, and capabilities people need to remember. That breadth is a strength, but it also creates a familiar problem: you know Kentik can do the thing you need, you vaguely remember being shown the workflow, but you can’t quite remember what it was called or where it lives. That gets even more pronounced for newer users, or for experienced users returning to a feature they haven’t used in a while.
Navigation Search is meant to make that whole experience less annoying. You can search for screens and features in natural language, jump around by keyboard shortcut, and use the search bar more like a modern command palette than a traditional portal search box. Just as importantly, search can surface the main knowledge base article for a feature right alongside the result, helping not only with navigation, but also with orientation once you get there. And because it respects RBAC and user-level access, people only see results they can actually use.
We like this update because it reflects a broader point about where we’re heading. AI should not only help after you’ve already found the right workflow. It should also help you get there in the first place.
What’s next
Kentik is becoming a more complete, more connected platform for teams that design, operate, and optimize modern networks. We’re continuing to bring NMS, flow, synthetics, and AI together more effectively, while also expanding capabilities like AI Advisor to make investigations faster, more consistent, and easier to operationalize.
We’re already seeing that value show up in the field. In Midco’s case, Kentik helped the team move from jumping between tools and spreadsheets to answering questions in seconds rather than hours, while Kentik AI Advisor accelerated time to insight and made troubleshooting more accessible across skill levels. In fact, they report lowering the mean time to insight by more than 60x.
Kentik should keep getting easier to use, easier to automate, and easier to operationalize, whether you’re a network engineer, a NOC leader, an SRE, a platform team, or a service provider operator. That’s where we’ve been pushing, and it’s where we’ll continue to push.
