Revision Date: 14 October 2022
This documentation has been created for software version 2022.3.1.2.
It is also valid for subsequent software versions as long as no new document version is shipped with the product.
Suite 105, 220 Davidson Ave, Somerset, NJ, USA 08873
Support:
For more information, visit
Copyright © 2022 whiz.ai All Rights Reserved.
Trademarks owned by whiz.ai
Disclaimer
No Warranties and Limitation of Liability
Every effort has been made to ensure that this document is an accurate representation of the steps to deploy WHIZ.AI platform. However, the development of the software is a continuous process. So, small inconsistencies may occur.
We would appreciate any feedback on this document
Send comments via email
Supported Browsers
WhizAI supports the following browsers:
- Google Chrome, v103 & above
- Mozilla Firefox, v103 & above
- Microsoft Edge, v103 & above (Windows); v102 & above (MacOS)
- Safari, v14 & above (MacOS)
What’s new
This document outlines the new product updates for the release version 2022.3.1.2. The following lists the new features and enhancements.
Pinboards, Cards, and Responses
Administration
Pinboards, Cards, and Responses
Responses - Supporting XLS export for Period Over Period (POP) and Comparison response
Now, you can download the XLS file for both POP and Comparison responses.
Period Over Period (POP) Response:
Comparison Response:
Limitations:
- When you reorder the columns on the response table and download the XLS file, the reordering of the columns is not captured in the XLS file.
- When you select or hide the columns from Columns (Refer to the following figure)
and then download the XLS file, the downloaded XLS file is not retain the selected or hidden columns.
- For iPad: When you download the XLS file and open it in the default application, the exported file displays only 4096 (Four Thousand Ninety-Six) records. However, all the records are displayed when you open the XLS file using Microsoft Excel.
Responses - Enhanced XLS and CSV file export on tabular response.
Now, for a tabular response that spans multiple pages, you can download the XLS and CSV files irrespective of the number of records in the response.
Limitations:
- When you reorder the columns on the response table and download the XLS file, the reordering of the columns is not captured in the XLS file.
- When you select or hide the values from Columns (Refer to the following figure)
and then download the XLS file, the selected or hidden values are not reflected in the XLS file.
- For iPad: When you download the XLS file with the default application, the exported file is displayed only 4096 (Four Thousand Ninety-Six) records. However, all the records are displayed when you open the XLS file using Microsoft Excel.
Responses - Distinguishing a custom narrative from auto generated narrative on the WhizAI explorer
Now you can identify a narrative as a system generated narrative (Auto) or a custom narrative (Custom).
To identify the narrative type, hover the cursor over the info icon on the narrative dialog box, and the narrative type is shown as a tooltip. For a system generated narrative tooltip is displayed as Auto and for custom narratives tooltip is displayed as Custom.
Insights Workbench - Supporting the best fit line on the scatter plot chart
Now, on the Key Drivers tab, you can see the best fit line on a scatter plot chart after performing the analysis to find out the top driving factors of your business metrics.
The best fit line indicates the relationship of different data points on a scatter plot chart.
To see the best fit line on the scatter plot chart please refer to the below example:
1. Go to the Insight Workbench and select the Key Drivers tab.
2. Select the FAS-Field Analytics from the Data Model dropdown list.
3. Select NBRx from the Metric dropdown list.
4. Add filters Period and Products and select last 52 weeks from the Period and Plabenil from the Products dropdown list.
5. Add factors Call Volume, Naive Volume, Customer Tier, and Primary Specialty.
6. Click Analyze.
You can see the best fit line on the scatter plot chart as shown in the following figure.
Insights Workbench - Supporting computed metrics on Anomalies and Key Drivers tab
Now, WhizAI has the ability to run Key Drivers analysis and Anomalies detection on computed metrics.
Note: Any metric other than the base metric is a computed metric. For example, it can be any computations that are supported by WhizAI such as Average, Growth, etc.
From the Insight Workbench, if you click the Metric dropdown list from the Key Drivers or Anomalies tab then you can see the list of computed metrics as shown in the following figures.
Computed Metrics on Key Drivers tab:
Introducing Business Data Descriptions
Business Data Descriptions are brief descriptions of dimensions and metrics. These descriptions provide basic and important information about the dimensions and metrics in the configured data model. Refer to the following table for examples of Business Data Descriptions.
Dimension / Metric | Description |
Product | Any medicine intended for human use, presented in its finished dosage form that is subject to control by pharmaceutical legislation (registered). |
NRx | A new prescription, defined as dispensed prescriptions given a new number by the pharmacy, not necessarily new therapy for the patient. |
TRx | Abbreviation of total prescriptions. TRx = NRx + Refills. After the first time a prescription is filled, when it is refilled, the pharmacy refers back to the previous drug ID number and notes this as a refill. |
NBRx | New brand prescriptions. |
Note: Business Data Descriptions are specific to the data model that is selected.
Business Data Descriptions are displayed on the user interface when you hover the cursor on a dimension/metric on the Info page or when you click on the Info option 
on the response. For more information on how to access data descriptions, refer to the following sections.
Accessing Business Data Descriptions on the Info page
The Info page displays a list of all the dimensions and metrics configured for the selected data model. You can access descriptions for these dimensions and metrics.
To access Business Data Descriptions:
- On the conversation box on the WhizAI explorer, click Info.
The following data model information page appears.
Accessing Business Data Descriptions from a response
You can access the descriptions of dimensions and metrics which are in the card context of the response, these descriptions help better comprehend the dimensions and metrics in the card context response.
To access Business Data Descriptions:
- On the NLQ response, click on the Info icon. New Information window appears. This window displays descriptions of metrics and dimension metadata and from the response.
To view definitions associated with dimension metadata, click Dimension.
For more information on how to use these options, refer to the Business Data Descriptions document: https://docs.whiz.ai/
Note: For optimal performance, it is advised that description of the dimension or a metric should contain 255 characters only.
Limitation:
- Business Data Descriptions are not supported for computations.
Viewing the NLQ on the card
Now, you can view the original query on the card that was used for generating the response. Therefore, if you want to regenerate the response using the original query, you can copy it from the Information dialog of the card and paste it into the context.
For example, when you ask WhizAI ‘Show me top growing regions’ then you get the following response.
When you click the ‘Info’ icon the Information dialog opens where you can see the Original Query as shown in the following figure.
You can click Copy and Paste in the context to regenerate the response.
NLQs - Using trimester and semester expressions in queries
Now, WhizAI supports Trimester over Trimester (ToT) and Semester over Semester (SoS) time expressions in NLQs. Semester and Trimester granularities are supoorted similar to the other granularities such as weekly, monthly, quarterly, yearly etc. You can include ‘by trimester’ or ‘by semester’ in NLQs to specify the time granularity in the response.
Trimester expression divides a period of twelve months into three periods of four months. If you include ToT expression in a query, WhizAI compares the given metric for the current four month period with the previous four month period, and gives you the absolute and percent change in the metric value.
Semester expression divides a period of twelve months into two periods of six months. If you include SoS expression in a query, WhizAI compares the given metric for the current six month period with the previous six month period, and gives you the absolute and percent change in the metric value.
For more information on the use of these expressions, refer to the following table.
Example NLQs | WhizAI response |
TRx in 2021 by trimester | TRx trend in 2021 by trimester. / TRx in each trimester in 2021 |
T1 2021 by months for TRx | TRx trend in the first trimester in 2021 by months. |
SoS | Compares TRx for the current six month period with the previous six month period |
NRx in 2021 SoS | Compares NRx for 2021 by semesters (H2 vs H1) |
ToT | Compares the given metric for the current four month period with the previous four month period |
ToT 2021 | Compares the given metric for 2021 by trimesters T2 vs T1 and T3 vs T2 |
TRx STD PoP | Compares TRx value for the period of current semester till date with the same period of previous semester |
TRx TTD PoP | Compares TRx value for the current trimester till date with the same period of previous trimester |
Example NLQs with the responses
Example 1: TRx in 2021 by trimester
Example 2: SoS
Administration
Script Editor - Editing existing scripts to create custom scripts
Now, you can add a new custom script by using any of the existing scripts. You can create a copy of an existing script (custom), and modify this copied script, and save it as a new custom script.
To add a new custom script by using existing script:
1. Go to the script that you want to copy, and from the Actions column, click the copy icon.
3. Script Editor window opens. From this new window, from the Information section, you can change script programming language and script usage. In the Editor section, modify the script and click Save.
New script gets added to the list as shown below.
Script Editor - Collapsing the Information section on the Script Editor user interface
Now, when you are adding a custom script, you can hide the information section on the Script Editor user interface and use the entire screen space to write/edit scripts.
To hide the Information section:
- Click and clear the Show/Hide Info Section option.
Now, you can use the entire screen space to write/edit scripts.
Data modeler - Viewing and accessing options under the Actions column on the Data Modeler user interface
Now, you can view multiple options under the ‘Actions’ column. (In the previous versions, these options were available only after cursor hover) For more information, refer to the following figures.
Data Modeler - Introducing Handbook on Data Load Configurations page
Now, on the Data Load Configurations page, you can click Handbook to get more information about the configuration settings required for data ingestion.
To access the Handbook:
Data Modeler - Adding Business Data Descriptions through data dictionary UI
Now, on the data dictionary UI, you can upload a CSV file containing descriptions of dimensions and metrics. These descriptions provide better understanding of the business terms and help you in understanding the business and asking relevant queries in WhizAI. Business data descriptions are displayed on the user interface when interacting with WhizAI.
Note: For more information on the Business Data Descriptions, refer to the WhizAI user manual.
To upload the CSV file:
- Go to the data dictionary page and click Upload CSV.
2. Select the Delimiter and choose the csv file that you want to upload.
3. Click Apply.
Note: For more information, refer to the WhizAI Admin manual- https://docs.whiz.ai/.
Data Modeler - Moving existing Airflow driven solution pipeline configurations to Data Modeler UI solution
Now, you can move existing Airflow driven solution pipeline configurations to Data Modeler and run the pipeline on the UI.
Prerequisite: Make sure you have WhizDMimport dag available and enabled on the Airflow v2.2.5.
To move existing pipeline configurations:
Go to WhizDMimport dag:
1. Click Trigger DAG 
.
2. Click Trigger DAG w/ config.
3. Provide a valid configuration JSON as shown in the following figure and click Trigger.
Note: For more information on this configuration, refer Understanding the WhizDMimport configuration JSON.
4. You can check the status of the DAG as shown in the following figure.
After a successful DAG run, all the pipelines will be available on the Modeler UI. You can Edit, Configure, and Run these imported pipelines on the Data Modele UI.
Understanding the WhizDMimport configuration JSON
Following table explains different fields in the WhiDMimport configuration.
JSON field | Description/Function |
code | This is Unique Identifier of your choice |
application_base_url | This is application url with the port number: |
region | It is AWS S3 region corresponding to access key |
username | Login credential: This is your login ID |
password | Login credential: This is your login password |
s3_access_key | AWS S3 access key, to be provided by customer |
s3_secret_key | AWS S3 secret key, to be provided by customer |
s3_bucket | Bucket to be used for data storage |
current_solution_dir | Directory used for existing ETLs by the Airflow driven DAGs. |
Narrative Templates - Introducing ‘click and select’ options to add function building blocks in custom narratives
Now, when adding a narrative template you can use ‘click and select’ options to add function building blocks in custom narratives. (Earlier you were required to write complex building blocks while adding narrative templates) For more information, refer to the following figures.
Before:
Now:
For more information on how to use these options, refer to the NLP guide: https://docs.whiz.ai/
Limitation:
- Narratives around Min. and Max. functions are not supported for NLQs having time series intent with other intents.
Known Issues
- When you run the Key Drivers analysis or Anomalies detection using a query that contains a computed metric, the computed metric is not displayed on the Key Drivers or Anomalies tab.
For example, if you ask ‘Show me average TRx for past six months’ and go to the Key Drivers or Anomalies tab, the computed metric TRx Average is not displayed on the Key Drivers or Anomalies tab, instead it displays the TRx.
- For NLQs with multiple time periods (for example, 4x4 , 8x8 and 13x13 by products), system generated (Auto) narrative is displayed for the first time period only (in this case 4 weeks).
- On pinboards, if you apply and save the filters (for example if you add Region, District, and Territory in the mentioned sequence) and then download the XLS file. The downloaded file does not display the filters in the same sequence in which they were saved on the pinboard.
- For a nested response, if you sort the dimension column/s and then download the XLS file, the sorting order is not retained in the downloaded XLS file.
Was this article helpful?
That’s Great!
Thank you for your feedback
Sorry! We couldn't be helpful
Thank you for your feedback
Feedback sent
We appreciate your effort and will try to fix the article