Easily Create Intent-Driven Dashboards Using Natural Language

cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 
Community Blogs
3 min read
L0 Member

prajamani_4-1769772989069.png

 

In the world of Network Security Operations, we often find ourselves "drowning in data but starving for insights." You know the drill: a critical ticket comes in, and suddenly you’re jumping between five different tabs, exporting logs to spreadsheets, and manually stitching together a story to find a root cause.

Traditional dashboards are great for answering yesterday’s general questions, but they often fail to answer today’s specific ones. That’s why we built AI Canvas.

 

From Static Views to Intent-Driven Insights

 

AI Canvas is a next-generation data exploration platform within Strata Cloud Manager that turns your natural language questions into interactive, high-fidelity visualizations. Instead of forcing you to adapt to a rigid UI, it allows you to build a workspace that matches your specific mission. Whether you're investigating a surge in risky app usage or preparing a regional threat report, you can now use natural language to define exactly what you want to see.

 

AI Canvas.png

 

Under the Hood: How AI Canvas Works

 

We know that for security professionals, "magic" isn't enough; you need to know how the data is being handled. The engine behind AI Canvas follows a structured, transparent pipeline:

 

  1. Translation: When you enter a prompt, we use LLMs to translate your natural language into a precise machine query (like SQL).
  2. Catalog Matching: The system identifies the correct data sources, such as Threat, Traffic, or ADEM logs from our cataloged data library that can actually answer your question.
  3. Raw Retrieval: We fetch the raw results directly from the logs/data sources, ensuring you are working with (almost) live, system-of-record data.
  4. Instant Visualization: The data is rendered into the most appropriate chart or table, which you can then export or add to your canvas.

 

Transparency is Not Optional

 

The biggest challenge with any AI system is the "black box" effect. We’ve solved this in AI Canvas by prioritizing transparency.

 

  • Catalog Visibility: You can see exactly which cataloged data sources are being queried, ensuring the AI isn't pulling from irrelevant datasets.
  • Query Validation: You have the ability to view the raw translated SQL query. This allows you to verify that the AI’s interpretation matches your actual intent before you rely on the insight.

    Transparency.png

 

Building Your Workspace

 

You don’t have to start from a blank slate. AI Canvas comes with pre-built canvases and widgets designed for common workflows. These serve as high-quality templates that you can quickly customize.

 

  • Multi-Turn Refinement: The conversation doesn't end after the first question. You can use multi-turn dialogue to refine your view, for example, "Now filter this by High Severity" or "Change this bar chart to a pie chart."
  • Intent-Driven Dashboards: Once you’ve built a widget that meets your needs, add it to a custom canvas. This isn't just a dashboard; it’s a purpose-built view for your specific strategic initiatives.

 

Beyond the Screen: Reporting and Collaboration

 

Insights are only useful if they can be shared. AI Canvas extends the life of your data beyond a single session:

 

  • Flexible Exports: Download any widget or canvas as a CSV for deeper analysis or a high-fidelity PDF for executive presentations.
  • Share and Schedule: You can share live, interactive canvas links with your team or schedule recurring reports to be delivered directly to stakeholders' inboxes, eliminating the "Friday morning scramble."

    Reporting.png

 

What’s Next?

 

AI Canvas is currently available for all Strata Cloud Manager Pro customers.

Stop searching and start solving. Log into Strata Cloud Manager today, open AI Canvas, and ask: "Show me the top 10 risky applications by session count over the last 7 days."

  • 1344 Views
  • 0 comments
  • 0 Likes
Register or Sign-in
Labels
Contributors