Use AI-generated insights to interpret a visualization

Chart insights provide AI-generated insights in the form of text summaries and key points to provide an easier interpretation of data in a visualization. The initial release of the Chart insights feature supports English only and is available only in Planning Analytics Workspace Cloud.

Important: The Chart insights feature is part of the Planning Analytics AI assistant add-on, which requires a separate license purchase; it is not included in a standard Planning Analytics Workspace license. To purchase a license, contact your IBM sales representative or go directly to your IBM account.

With Chart insights, depending on the visualization type and the available data, you can review important information that helps you interpret a visualization. The Summary tab variably provides an overview, identification of statistics and trends, conclusion, and analysis relevant to the contextual information in the chart. The Key points tab provides a summary of the visualization, trend conclusion, brief listings of the key statistics, notable observations, and insights.

AI-generated insights for a chart

You can view details of the AI model and an overview of how it works in this feature by clicking the AI label.

Generating insights

To generate AI insights for a visualization:

  1. Select the visualization that you want the analysis for and click Chart insights on the toolbar.
    Chart insights icon displays in the toolbar

    AI starts generating a text summary of the data in the selected series.

  2. After the summary is complete, you can click the Key points tab for insights and notable observations.
    Note: You cannot navigate between the Summary and Key points tabs during text generation. You also cannot regenerate or copy the response until after generation is complete.
  3. If your visualization has more than one series, AI insights are generated for the first series in the visualization. For example, if your visualization includes a series that includes individual months, the insight is generated for January. To generate insights for a different series, click the Select a chart series to analyze menu and select a different series.
    Series menu
  4. Optionally, click Regenerate response Regenerate response iconto regenerate the summary and key points. The response will vary slightly each time you regenerate.

Chart insights supports the following charts for analysis:

  • Area
  • Bar
  • Box plot
  • Dual axes column
  • Dual axes line
  • Line
  • Line and column
  • Pie
  • Point
  • Radar
  • Radial
  • Stacked bar
  • Stacked column
  • Waterfall
  • Word cloud

AI explained

The Chart insights feature uses granite-34b-code-instruct from the Granite family of IBM foundation models to generate summaries.

The Granite models are decoder-only models that can efficiently predict and generate language. These models were built with trusted data that has the following characteristics:

  • Sourced from quality data sets in domains such as finance (SEC Filings), law (Free Law), technology (Stack Exchange), science (arXiv, DeepMind Mathematics), literature (Project Gutenberg (PG-19)), and more.
  • Compliant with rigorous IBM data clearance and governance standards.
  • Scrubbed of hate, abuse, and profanity, data duplication, and blocklisted URLs, among other things.
Note: IBM is committed to building AI that is open, trusted, targeted, and empowering. For more information about contractual protections that are related to IBM indemnification, see the IBM Client Relationship Agreement and IBM watsonx.ai service description.

To generate a summary and insights, the AI model takes key inputs such as the chart data and calculations from the database. The AI model can analyze information from the selected chart series and calculated statistics to describe the chart, data points, and any trends presented in the data. A text-based summary is then generated by the model along with key insights for the selected chart series.

Tips and best practices

While working with Chart insights, here are few things to ensure that might lead the AI model to generate better insights:

  • Chart has data and is not empty
  • Chart data is in context of the cube
  • Cube, dimension, and member names are descriptive and accurately represent the data
  • There are no named outliers for dimensions that might confuse the AI model