Configuring CADDI - Creating Chat Bots

Introduction 

Conversational AI for Data-Driven Insights (CADDI) makes it effortless to interact with your data, understand it, and extract reliable insights by using plain English. This article explains how to configure a CADDI chatbot.

Overview

  • Create a chatbot
    • Select tables and columns
    • AI model configuration
  • Train bot
  • Detailed Question
    • Logs Generation & SQL Query
    • Different Generated Views
    • Saving the Question as Context
  • Share a Concept.
  • Data managing operations

Detailed Steps

1. Create a Chatbot

Chatbots can be found on the source detail page. Any source that has chat enabled will have the Chatbot tab. Chat Access can be given to any source through the Module Settings. For more information, click on User Role and Module Settings

The Chatbot Tab can be found on the detail page of the source for which Chat is enabled. 

datasource.png

The "Chatbots" tab can be seen next to the "Settings" Tab. On clicking on Chatbots, a new Chatbot can be created:
createchatbot.png

 

Upon the creation of a new chatbot, users are directed to the Definition Tab, where they can configure settings.

 

1.1 - Select Table and Columns

Within the data source, multiple tables may exist. Users have the option to either incorporate all tables into the chatbot or to select which tables will be available.


 

The added table shows up in "Tables For Chat". Users can also preview the Table to see the data inside it:


viewtable.png

 

Users can access individual columns and their profiling by selecting the "Manage Fields" option. Within this interface, users have the ability to choose or deselect columns from any table.


managefields.png

 

Once the required tables have been added, users can save their changes by clicking on the "Done" and "Save" buttons.

 


1.2 - AI Model Configuration

Users can select their desired AI Model for the Chatbots. GPT-3.5 Model will be selected by default but different AI Model can be select. To change the AI Model, click on the 'Edit' button on Definition tab.

 

Click on the Chat Model dropdown under the Configuration heading. Click on it to view all available AI models, then choose the one you'd like to use.

After selecting the desired Model click on 'Done' button and than click 'Save'

 

Any settings configured at the chatbot level will apply not only to the Train bot but also it will reflect to all the shared chats at CADDI.

 

2. Train Bot

After choosing tables and saving, users can find the "Train Bot" tab. Clicking it, they'll see the "Generate Questions" button and a chat bar for typing questions.


generate questions.png

Clicking on "Generate Questions" displays a set of sample questions. Users can utilize these questions for analysis purposes.

 

3. Detailed Question

A detailed question is the one for which:

  1.  SQL Code, Logs and explanation are generated 
  2.  Visualized Grid, Charts, and Pivot Information are generated
  3.  Context can be set up and Profiling Information is provided
  4.  Positive or Negative feedback can be given for further improvement (Training)

3.1 Logs Generation & SQL Query

In simpler terms, logs contain information about when a question is asked and what is sent to the backend in text form. To view logs, users can enable the logs by clicking 'Show Log' icon at the end of question block.

After enabling the logs, the corresponding logs will be displayed. Additionally, the SQL query and Explanation generated for that question will be printed below the logs.

The Log details, SQL code and the Explanation can be hidden by clicking the 'Hide Code', 'Hide Log' and 'Hide Explanation'

3.2- Different Generated Views

CADDI generates three different views:

The Grid View shows the data in tabular form of the columns that are fetched from the SQL query:

 

3.2.1 Default Answer View

By default, when a question is asked, Chart view is shown with the response; however, users can choose to set the default view to Grid View. To make this change, go to the 'My Settings' and select Chat Preference tab, select the Grid option, and save your selection.

It's important to note that this updated setting will not affect previously asked questions; only new questions asked after the settings have been applied will be impacted. These bot settings will also apply to the CADDI conversations that were linked to this bot.

 

Along with the 3 different views there will be a gear icon of Settings which is present when this is clicked settings tab opens up with further three options:

  1. General 
  2. Chart (Learn more in Chart Settings in CADDI)
  3. Field (Explained in Profiling & Filtering below)

The General tab contains the toggle button of "Human readable format". By default, this option is turned off:

 

Enabling this option improves the readability of data, especially when dealing with large values, such as thousands or millions. It presents the information in a more readable format for easier comprehension.


 

3.4- Saving the Question as Context

The question asked, along with any additional context or specific filters applied to the grid, can be added using the Context icon located at the top right of the question.


The added context is visible on the top left of the chat bar below. Users have the flexibility to include multiple contexts from different datasets. If a context is no longer needed, it can be removed by selecting "Clear Context."

In the future, if a dataset-related question is set as context, it will automatically be included in later questions.first_part.png

added_context.png

 

4- Share a Concept

Similar to providing feedback, users can share a concept for a specific thing they want to include in upcoming questions. The "Share a Concept" option is available at the plus button next to the chat bar.
shareaconcept.png

save_concept.png

 

added_concept.png

 

When a question related to this table is asked next time, the added concept is included in the response if needed:


onquestionconcept.png

5- Data managing operations

On the top right of the Chat Bot, some basic operations can be pretty handy for the users. There are two filters which are:

  1. Asked By
  2. Status

The "Asked By" option allows filtering questions based on different users in the chatbot. This filter displays all the various users who have asked questions in the chatbot.

There is another filter called "Status" that categorizes questions based on different question types. The available filter values are:

  1. All 
  2. To Review 
  3. Failed to Generate (Questions that gave error)
  4. Approved (Questions given positive feedback)
  5. Rejected (Questions given negative feedback)
  6. Concepts 

 

With the two filters, there are three dots that provide users with the option to "Refresh All," refreshing all the questions. Additionally, users can choose to "Expand/Collapse All," which expands or collapses all the questions present in the chatbot.

Users have the option to duplicate or delete the chatbot. If a user wishes to work on the same chatbot without affecting the original, they can duplicate it. Alternatively, if the user decides to remove the chatbot entirely, the delete option is available alongside the duplicate button.


 

Below the chatbot name, additional basic information about the source and chatbot is available. This includes details such as the number of questions asked, associated tables, and chat exclusion criteria, all of which are specified in the Module Settings for this source. Module Settings Article.


exclusion critera.png

Additional References: 

CADDI- Generating Answers Based on Different Tools

Manage access to the ChatBot

CADDI- Profiling & Filtering

CADDI - Feedback

Have more questions? Submit a request

0 Comments

Please sign in to leave a comment.
Powered by Zendesk