CSV uploads in beta!

CSVs can now be uploaded in workbooks, making it easier to bring in external data without adjusting existing pipelines. For example, you may want to upload data that isn’t typically found in a data warehouse, such as quota or goal data.

Check out the docs for all of the details. If you’re interested in taking the CSV beta for a spin, reach out to Omni support to get started!

Folders, now with levels

You asked, we delivered: folders can now contain subfolders. Up to seven levels of subfolders, in fact.

As folders can be used to control how content is accessed, we recommend checking out the updated docs to understand how subfolders and access roles work together.

More AI updates: Bigger, better, Blobbier

Blobby, everyone’s favorite AI helper, can now do more than just help you write queries: Blobby can summarize insights and create filters.

AI-powered summarization

With Blobby’s summarization functionality, you can ask to identify anomalies in the data, identify insights into trends, what next steps you should take, and more. You can enter prompts directly into the query helper:

Or create a new AI summary visualization to include the summary in a dashboard tile. You can also provide additional context to guide the AI:

The AI summary features will be generally available soon, but feel free to ping Omni support if you’d like us to turn them on for you.

Magic (okay, natural language) filters now in beta

You can now create natural language filters right in the workbook by clicking the funnel icon next to the Options button. Describe your filter, such as Show me issues created in Slack for the past 7 days, and hit Enter:

And watch the magic happen as the filter is added to the workbook:

If you’re interested in trying this feature, reach out to Omni support.

Fine-tuning model performance 

We’re always working hard to make Omni more performant. In the past few weeks, we added the following model-level settings:

• filter_suggestions_must_obey_policy, which controls how long filter suggestions are retained

• warehouse_override, which allows you to override the connection being used for a specific workbook. This can be useful if you need to make a specific dashboard more performant, such as providing more compute for querying. Note: This setting may only be used in workbooks, meaning it can’t be promoted to the shared model, and only applies to Snowflake and Databricks connections.

User attributes, now *on* the users page!

A user’s user attributes can now be accessed on their user page with the Attributes tab:

Readable data labels in charts

Gone are the days of overlapping, jumbled data labels in charts! We recently released an algorithm that better displays data labels and results in readable labels every time.

An example of what data labels were before:

And what they are now:

If you need to utilize the denser data label approach, you can do so by enabling Show All Labels in the chart’s axis settings.

Lots and lots of API enhancements

Our library of APIs have gotten a lot of attention recently, including:

• New endpoints that allow you to retrieve lists of documents and folders. These endpoints also support pagination, filtering, and sorting.

• The ability to run queries as a specific user by using the new userId parameter in the Run queries endpoint

• The List users endpoint now supports filtering embed users using the embedExternalId parameter

New reading material in the docs

We’ve also released a bunch of new documentation!

• Several new copyable Markdown viz examples:

  • Card with strip plot

  • Card with target candlesticks

  • Gauge

  • Gauge with context

• A guide to dynamically switching database environments using user attributes

• A guide to help you migrate dashboards from one instance to another using the content migration APIs

• GitLab and Azure DevOps-flavored setup instructions for the git integration

Honorable mentions

And last but not least:

• The Summary Value visualization option has been deprecated. Existing charts using this visualization type will continue to function, but going forward:

  • If you select a different option for an existing Summary Value chart, you won’t be able to go back

  • For new visualizations, Omni will use the KPI option as the best default

• A new 2D projection option for point map visualizations

• Document identifiers can now be up to 48 characters

• You can now set the opacity for a color in the chart color picker: