Description
This video is a recording of Kevin Kamel’s presentation at the AI Networking Summit (ONUG) Spring 2024. Kevin introduced Selector as a platform to transform operational data into actionable insight and discussed three critical operations challenges Selector solves with AI.
Video length: 8:55
Speakers: Kevin Kamel, VP of Product Management at Selector
Transcript
[Music]
All right, thanks, folks. I’ll just jump right in. So, what is Selector? Selector is a platform that transforms your operational data into actionable insights. We come into your network and IT environment, establish what’s normal, and then notify you when there’s a deviation so that you can respond.
I’m Kevin Kamel, Head of Product at Selector. We focus on three key areas:
- Collection and Management of Heterogeneous Data
We capture metrics, logs, and events from all your equipment, primarily on the network side. Whether you have a multi vendor setup, we can collect data without the need for vendor-specific software. This data is then warehoused within our stack, where we apply machine learning and AI to automatically surface insights. - Automated Insights and Correlations
Today, operators manually correlate data from different platforms like metrics and logs, often using multiple displays in a NOC. Our platform automates this process, delivering correlations so you can take action without the manual legwork. - Addressing Siloed Collaboration
In many NOCs, not everyone has access to all the tools or data they need. With Selector, we’ve integrated a generative AI directly into our platform, allowing conversational interrogation of your Telemetry data. Whether you’re using Slack, Teams, or another collaboration tool, our platform ensures that your entire team can work together to resolve incidents.
Now, let’s talk about data collection. Selector goes directly to the source—network devices, infrastructure, applications—and gathers Telemetry data. We also support proprietary data sources with a no-code, low-code integration platform that lets us quickly build custom integrations.
Once we’ve collected the data, we apply machine learning to establish baselines and detect deviations. These deviations are logged as events, which we store for later correlation. We also tag events with contextual data, such as router information or BGP session details, which we can use for further analysis.
We then perform correlations using temporal, contextual, and topological data to group related events into incidents. Finally, we apply causal ML to identify the root cause and related events, allowing you to focus on the most critical issues.
Operators want smart alerts that provide root causes and related activities. With this information, they can remediate issues and understand the broader impact, such as which customers were affected.
In recap, Selector’s platform offers auto-baselining and thresholding for metrics and logs, event extraction, named entity recognition, and a conversational LLM that lets you query your Telemetry data directly within Slack or other collaboration tools. No need to learn complex languages—anyone can use it.
The beauty of this is that even my six-year-old could ask, “What are the events related to this?” and get an answer within Slack. It’s an entirely new way to engage with your Telemetry data. If you’d like a demo, our booth is in the back right of the room. We’d be happy to show you more.