Description
Selector Copilot allows users to conversationally interrogate their network telemetry by leveraging a natural language interface to retrieve and render analyzed insights from the Selector Analytics platform. This video shows how Copilot interprets queries, translates them into actionable insights, and provides contextual responses through collaboration tools like Slack and Teams. The demo includes a real-world example of Copilot in action, aiding operators with network management tasks.
Video length: 1:52
Transcript
[Music]
Narrator: Copilot leverages a natural language interface to retrieve and render analyzed insights from the Selector Analytics Platform and is accessible through preferred collaboration software like Slack and Teams. Internally, Copilot leverages a combination of an integrated LLM to interpret natural language queries and generative AI to summarize the results in a human readable way. To quickly explain how this works, Selector LLM is trained with deployment specific vocabulary. The training data comes from the user intent extracted from the dashboard widgets and aliases created by the users. Natural language queries from users are translated into Selector Query Language by using this model. The query results are then fed into the LLM for summarization and contextual responses. The responses are augmented with any recommended actions based on the preconfigured workflows. But rather than talk too much about the internal, let’s see a real world example of how Copilot is helping network operators today.
[Music]