cancel
Showing results for 
Search instead for 
Did you mean: 
ChrisS
Sisense Team Member
Sisense Team Member

Infusion Apps: Ask Me NLQ

What’s Ask Me 

Sisense’s Infusion Apps enable their users to pull their data directly into business applications, including Slack, MS Teams, Office 365, and Google Suite.  Curated data, Views, are accessible with the same login credentials they already use to access business insights in your applications.  The data delivered is automatically filtered to your data security rules for each user, so the data is completely governed.  My favorite  Infusion Apps feature is being able to ask for the data to be filtered the way I want using Natural Language Queries (NLQ).  This post is about providing a primer for how users can use the Ask Me feature in Infusion Apps to get just the data they want in reporting and visualizations.

What I want you to get out of it

  1. A Quick Start to Infusion Apps’ Natural Language Querying of Data Views.
  2. A broader familiarity with Infusion Apps can bring consistency to teams using data previously subject to all the vagaries that arise from extracts pulled at different times by different users.  Everyone can sync on the latest data by hitting the Refresh Data button!

Using Ask Me

1. The Basics:


The NLQ in Infusion Apps, Ask Me, is not the same as Simply Ask in dashboards.  Ask Me simplifies the querying process by allowing users to employ jargon they are comfortable with while view names, column names, and literal values must be spelled correctly. Each request has to reference a View by name, including spaces.  Ask Me isn’t case sensitive.  For example, given Views named “Orders” and “West Orders”, Ask Me will return the Orders and West Orders lists for any of the following commands:

  • Show West Orders
  • Show orders
  • I want oRders
  • Show me ORDERS
  • What are orders

The basics for getting started include how to ask for help, which is simply, “help”, and get a list of views, just type “views”.  It’s going to be that easy to work with…

 

ChrisS_0-1682004967835.png

 

Sisense Documentation: https://docs.sisense.com/main/SisenseLinux/using-ask-me.htm?Highlight=Ask%20Me

 2. Working With Views: 


Views are the data sources we can query using Ask Me.  They are practically as easy to create as querying them with Ask Me in your applications. Views are simply tables created in Sisense and published to users.  Flat tables are an ideal format for user self-service as a best practice.  We’re only covering tables and their creation briefly, where it informs how we can filter data from the Views with Ask Me.  

Designers create tables and can specify things like whether fields are visible, what aggregation should be used for measures, primary date columns, and other View configurations.  A particularly useful part of the process is setting up filters for the data in the table because it enables us to use them in the conditions of an NL query. Fields have to be defined as filters to be able to use them in your conditions.  Notice in the View below that there are more Columns than Filters, and only designated filter fields will work in Ask Me.

 

ChrisS_1-1682005080229.png

Columns & Filters in the Presentation Data View via Slack

 

Designers can also designate aliases for the filter fields, which is super handy for enabling non-data techies to use their own vernacular for dimensions and measures.  For example, the actual data column name might be “Value”, while the sales team thinks of it as “Sales”, CROs and Finance think of it as “Revenue”, etc.  The designer can set a field’s display name, but they can choose only 1. For supporting users with different perspectives you can add aliases for Filter fields that make querying by their team or department easier and faster.  Given a column named, “Opportunity Name”, we can add aliases as below, and users will be grateful not to have to type out the lengthy column name every time.

 

ChrisS_2-1682005125960.png

 

How do you know what columns are available and which filters can be used?  It’s simple, every View advertises the Columns and Filters users have for Ask Me.

3. Using Ask Me

You probably noticed there are some good examples of NL queries in the Help output above.  An important takeaway is that you can include multiple filter clauses separated by “and”.  It’s helpful to remember that the data is in a flattened table. The objective with Ask Me is to select the data easily.  Shaping the data and creating tables and charts is work accomplished in the Infusion Apps, e.g. MS Excel or Google Slides.

Slack & Teams

Ask Me is essential working with  Slack and MS Teams.  A frequent task I perform is tracking the performance of teams and their members.  Here’s our Orders view in Slack

 

ChrisS_3-1682005176386.pngA View in Slack

 

Knowing that Ben, David, and Sam are the Enterprise AEs, I ask the following in NLQ:

show orders last year whose ae = ‘ben,david,sam’  and sales > 15k

Let’s break down the query:

  • “Last Year” refers to 2022, 
  • ‘ae’ is an alias for Sales Rep, 
  • ‘sales’ is an alias for Amount  
  • “= ‘ben,david,same’” can be read as “contains Ben or David or Sam”.  

These aliases were added to the filters for the view.  Here are the results:

 

ChrisS_4-1682005226232.png

 

Of course, there are more pages in the output that I can click through, and even review additional details or other columns not listed for querying by clicking on the ellipsis. 

It’s easy to share this data set as a bookmark with any peers who have access to the same data.

 

ChrisS_5-1682005263451.png

Slack Bookmark

Notice that Sisense includes the query with both the result set and the bookmark.  Infusion Apps are able to share bookmarks across the other applications supported.  Bookmarks aren’t Global objects, though, so not sharable themselves, but the NLQ behind them is sharable with others.  Let’s open the bookmark created in Slack with MS Teams:

 

ChrisS_6-1682005342768.png

MS Teams

So all the Infusion Apps work with each other. Slack and Teams can also summarize and visualize the data for you by clicking on the Analyze Data button.

 

ChrisS_7-1682005379803.png

Analysis in MS Teams w/Summary & Chart

Google Slides Example

Ask Me is available in all the Infusion Apps.  Here is a Google Slides example.  We’re going to create a Summary & Chart for a slide in one of my Business Review slides.  We access the Ask Me section, compose our NLQ request with the Columns and Filter fields for the View, and Sisense inserts the objects directly into my slide.

 

ChrisS_8-1682005418686.png

Ask Me in the Infusion Analytics Extension

 

We ask for the orders of Mike, Molly, and Rahul, our sales reps.  Then we can click on a button for a Summary, Chart, or Table to insert it into the slide. When we select Chart, the chart builder is invoked so we can do what we normally do in this workflow, which now happens all in the Infusion App, instead of me having become the human middleware to fetch an extract.

 

ChrisS_9-1682005456727.png

Google Slides Chart Builder

 

Below is the finished slide.  

 

ChrisS_10-1682005493999.png

Finished Google Slide using Ask Me

Charts built using Ask Me results aren’t able to be refreshed, but summaries, charts, and tables created using the Views can be refreshed at the push of a button individually or en masse using the Refresh All link at the bottom of the Extension panel.

 

4. Summary

Ask Me is the easy way to talk to your data, especially when using Slack or Teams in a meeting.  It’s great to have answers in real-time or provide a productive way to offer governed data to users in their existing workflows.

Call To Action

Extend your Governed Data’s reach into the workflows not designed around dashboards, and recoup as much of the “being middleware” time as you can!

1 Comment
Contributors
Community Toolbox

Recommended quick links to assist you in optimizing your community experience:

Developers Group:

Product Feedback Forum:

Need additional support?:

Submit a Support Request

The Legal Stuff

Have a question about the Sisense Community?

Email [email protected]

Share this page: