The Natural Language Querying (NLQ) tool lets users ask analytical questions in plain English, and get a response as a result-set presented in the most appropriate visualization. Creating the response makes use of an LLM Engine, selected by your administrator.
In effect, NLQ offers users a way to query data quickly and intuitively, without the need to navigate the details and elements of data models like dimensions, measures, or elements trees; without having to learn or understand how to use and place 'chips' in the drop zones; or even how to select and configure an appropriate visualization. Instead, Pyramid interprets the questions and automatically finds the relevant hierarchies, measures, and elements, places them into the appropriate drop zones and selects the optimal visualization for the query.
Note: NLQ is available with an Enterprise edition license only. If you are using the "Pyramid Internal" Engine, the Chatbot and NLQ functionality works with ENGLISH only.
Note: Pyramid's NLQ operates directly on ANY supported database (both SQL and MDX).
Using NLQs in Discovery
NLQ is one of five different Discover modalities that are available to users based on licensing, user profiles, and system activated options. Arguably, NLQ represents the simplest way to perform data discovery in Pyramid because it does not require extensive training for users - relying instead on advanced AI functionality to interpret questions in the context of any presented data model and derive both the asked question and find an appropriate result.
NLQ can be used to generate a data discovery; this is a single-step process involving asking a question. NLQ can also be used to make changes to the query, and even interact with it using the NLQ Chatbot - where you can ask questions cumulatively in order to continue building and manipulating the query,
Accessing NLQ features
Before you can use the NLQ, it must be enabled from the AI Settings in the Admin console.
Once enabled, NLQ is available to Pro licensed users from within both the Discover Pro and Discover Lite tools. It can also be used when adding new content directly into Present and Publish; and from the 'analyze further' option and Chatbot in Present dashboards when viewed at runtime.
- To learn how to access and launch queries using Natural Language text bars (Discover only), see Accessing NLQ in Discover
- To learn how to access the Chatbot in Discover, Publish, or Present, see Access the Chatbot at Design Time
Asking Natural Language Questions
The NLQ engine is designed to accept plain English language questions that drive data queries and create visualizations of the results. At its most basic, the questions need to reference elements from the data set being targeted for analysis. However, there are numerous "smart" aspects to the engine that can both understand and impute analytic computations from various types of phraseology, word variations (aka word stemming) and key question structures that drive analytic concepts.
Understanding these nuances allows you to drive smarter more intelligent questions and analysis.
- For more information about writing natural language queries, see Writing Natural Language Queries
- For more information about using the Chatbot to enter your NLQs, see Chatbot Functions
Example: Build a Visual using the Chatbot
- For a step-by-step guide to using the Chatbot to create a Discovery, see Example: Build a Visual using the Chatbot
Managing NLQ Discoveries
Since NLQ discoveries are, in effect, normal Discovery reports created using the AI-driven tools, there is no special framework for managing them specifically. It is, therefore, the case that, once you have finished building and editing your visual, you can manage your content as you would any Discovery.
- For more detail about managing the Discovery reports that you created using AI, see Managing NLQ Discoveries