The field of business intelligence (BI) is changing quickly these days as customers utilize strong, generative AI-driven data analysis tools to extract ever-more value from their own information. Google Cloud is releasing LookML Assistant and Visualization Assistant, two crucial Gemini features in Looker, to enable you create and interact throughout your company with intuitive AI-driven insights, and to assist speed up this change. These developments streamline the organization and distribution of data for analysts and developers while reimagining how they can engage with modeling and visualization tools.
Unveiled at Google Cloud Next, in April, Looker’s Gemini will alter how you derive insights from the most consequential data available to you. You can help everyone in your company by leveraging natural language and Google‘s AI to explore, collaborate, and make decisions more quickly among your team members.
Use the LookML Assistant to write LookML more quickly
Looker’s version-controlled semantic layer, or LookML Assistant for short, serves as an abstraction layer between your data and business users, giving them a dependable and consistent means of accessing and analyzing information. This makes data governance easier to manage and promotes data integrity throughout the company. LookML Assistant streamlines the code-creation process and offers recommendations and direction to help you get started with LookML and expedite your workflow. LookML Assistant can be used to generate measurements, dimension groups, dimensions, and other code recommendations.
Looker’s Gemini offers generative AI powered support to ease your data processing.Using Looker’s Gemini, the Looker IDE’s Help me code screen uses written instructions to produce LookML code suggestions. In your LookML Assistant, Gemini can be used to construct dimensions, dimension groups, and measurements.
To include LookML into your Looker project using Gemini, take the following actions:
- You need enable Development Mode on your Looker instance.
- Launch the Looker IDE and open your project.
- To put LookML into a file, open the LookML view file using the IDE file browser.
- Using the side panel selection, choose the Help me code icon.
- Click to highlight a line in your LookML view file while the Help me code panel is displayed. Gemini offers relevant options, such Create a dimension or Create a measure, to assist you based on the type of LookML file you are working with and where your cursor is located within the file.
- From the Help me code panel, pick one of the following:
- Form a group of dimensions.
- Construct a dimension
- Make a measurement
- Other code suggestion: If you’d like to experiment with other LookML elements, you can use this option. Since Gemini is still in its early stages of development, make sure to verify and test all of the output before implementing it.
- Use natural language when describing the dimension, dimension group, or measure that you wish to construct in the Help me code panel text section. For assistance, go to the parts on this page that include Sample prompts and Tips for using Gemini in the Looker IDE.
- To submit your request to Gemini, click the Submit icon or hit Enter. Gemini will reply with a code recommendation.
- With the provided code, the following actions are possible:
- To see a preview of the proposed LookML in your file, hover your cursor over the Insert button. Before adding the LookML to the file, you can shift your cursor to a different line to preview it in a different place.
- To make manual changes to the proposed LookML, click the Edit button.
- To put the LookML into your file at the point where your cursor is, click the put button.
- Please select the thumbs up or thumbs down (This response was not what expected) to indicate your opinion on the provided code.
- To start anew and input a new Gemini prompt, click the New code proposal link.
- In your LookML file, click Save Changes.
- Check the LookML by carrying out these actions:
- Check the LookML with the LookML Validator.
- Take a LookML test in an Explore.
Create bespoke visualizations with natural language
Additionally, by adjusting the underlying JSON code directly, you may fine-tune the look and functionality of your visualizations using Looker’s Chart Config Editor. With the considerable customization possibilities this tool offers beyond the usual visualization settings, you may make charts and graphs that are both more visually appealing and instructive.
With the release of Visualization Assistant in preview today, google cloud extending this capability and making it simpler than ever to add adjustments to Looker’s pre-made visuals. To personalize your visualizations, ask Gemini in Looker using natural language, rather than scouring the docs for the appropriate JSON options.
Create a visualization in Looker Explore and select an out-of-the-box chart type, like a column, bar, or line, to begin using LookML Assistant After that, select the Visualization Assistant button from the options bar for visualization. Looker’s Gemini will inquire as to what you would like customized. When you provide a question or choose one of the pre-written suggestions, Gemini generates the customized code. Apply the parameters and your new visualization will go live if you like what you see.
Looker’s Gemini feature lets you communicate with your data to make it more interactive and expedites daily data engagements.
