Register

Click the blue GladePay button below, you would be redirected to GladePay website. On the top-right corner of the website, 'Sign Up' and select "GladePay Social" as the "Services". Alternatively, speak with the ChatBot on the bottom left corner of GladePay website and complete signup from there.
Use login credentials sent to your provided email to set your ChatBot service up

Gladepay

Enter your desired ChatBot details

Step One

Write basic information about your ChatBot

Enter ChatBot Name, Description, and Select the language you intent to use in training your ChatBot. You should add additional settings for channels(Facebook, Twitter, WhatsApp, Web, Slack, etc), theme color. Please note that the channels available to you depends on the package you subscribed for.

Your ChatBot Details
chatbot details

Create chatbot

Step Two

Create new ChatBot

Click on "Add new ChatBot". This leads you to select a subscription plan

Choose a subscription package and billing cycle. That is, 'Standard', 'Premium' or 'Enterprise'

Select subscription plan and make payment
elegant icons
Next

elegant icons
Next

elegant icons

Make Payment

Step Three

Make payment through GladePay checkout

Enter your billing details and make payment

Make payment
elegant icons

Customize ChatBot

Step Four

You can now train your new ChatBot and customize to suit your business needs

Open your chatbot to train and define the tasks you wish to accomplish with it

Open the ChatBot
chatbot details

ChatBot Training Components

  • Natural Language Understanding

    Natural Language Understanding is the component used to train the chatbot on what it should understand when people interact with it using human languages such as English

  • Intents

    Intents mean the actual interpretation that the chatbot makes of what a user says. For instance, a user says "Good morning"; the ChatBot reads this as a 'greeting' Intent. You are the one to specify the range of intents that you want the ChatBot to interpret.

  • Entities

    Entities are keywords that help the ChatBot understand an Intent better. For instance, 'Good morning', even though the Intent is greeting, the keyword 'morning' can be used to let ChatBot know that the greeting time is in the morning.

Open your chatbot to train and define the tasks you wish to accomplish with it.
Ensure you do not use the same name for two different intents or entities

Training Components
nlu details
Add Intent
add intent
Choose a name and describe what intent does
add intent add intent
Add Entity
add intent
Choose a name and describe what intent the entity is attached to
add intent add intent

ChatBot Training

  • Train

    This is where the ChatBot gets trained with the sentences users would likely use when talking with the ChatBot

  • Expressions

    Possible user sentences or phrases are entered here. It is advised that you start training with single words and phrases before moving on to complex sentences.
    Note that only one sentence can be used to train at a time.

  • Specifying Intent

    Select an existing intent that you want the user's sentence to be interpreted as. You can also create new Intents from here by simply typing the desired intent name.

  • Entity category

    Select or create a new Entity that should be used for this sentence.
    You can create a new Entity by simply typing the desired entity name

  • Entity value

    Select the keyword to be picked from the sentence for training.
    Note that you can train with the same sentence many times in order to pick different entity values

  • Full and Partial Training

    The ChatBot may be trained with 'Expression + Intent + Entity + Entity Value' or with 'Expression + Intent' only. This is referred to as Full and Partial Training respectively

Open your chatbot to train and define the tasks you wish to accomplish with it.

Training Fields
nlu details
Sample Training
add intent
Trained Sample
add intent

ChatBot Responses

  • Flows

    When a user's statement is rightly understood by the ChatBot (i.e., when the ChatBot assigns an Intent to a particular user sentence), Flows are used to tell the ChatBot what kind of response it should give back or action it should take.

  • Messages, Questions and API requests

    The ChatBot may send a message back to the user, and/or, ask a follow-up question to get more information needed to accomplish a task, and/or send data to an API

Click an Intent to define its Flow
nlu details

Flow Tasks

There are three kinds of Flows. Message, Question or Webhook (or API service). These three define the the kind of action that the ChatBot takes when it receives input from a user.

nlu details

Add a new response type (message, question or API request) when this Intent is met.

nlu details

Message Response

This is used to specify messages that the ChatBot responds to the user with when an Intent is matched and certain conditions are met. Conditions to be met are defined within the 'Filter' tab. Filters

nlu details
    Message Task Terminologies
  • Task Name

    Specify a unique name for each task. This name is used in filters when setting conditions that must be met before that response is given to the user.

  • Message

    This is the response the ChatBot should give when the Intent conditions are met.

  • Variation or Alternate Message

    These are different ways of presenting a response with the same meaning. This is done so that the ChatBot does not respond with the same sentence every other time. For instance, the ChatBot can either say "Hello, welcome today" or "Hi there, you are welcome".

nlu details
    Filter Terminologies
  • Filter Tab

    From the filter tab, all conditions that must be before the response is sent to user are specified here.

    NOTE: FILTERS MEAN THE SAME THING FOR ALL KINDS OF RESPONSE TYPE (i.e., message, question, API request, image, etc... response types)

  • Filter Field

    This represents an extracted value from a user's input. It could be the user's name/entity keyword/session data/etc...

  • Value

    This represents ChatBot set rules (i.e., ChatBot rules) or values that the chatbot expects user input to be.

  • Conditions

    These are used to compare what the ChatBot got from the user (i.e., user input) and ChatBot set rules (i.e., ChatBot rules). It may compare Field with Value to see if they are equal/not equal/contain similar information/greater than/less than/etc...
    For instance, if an entity is picked from a user as "colour" but the chatbot expect a Value pf "height", the condition "Equals" would fail, hence this message would not be given as a response.

  • AND Group

    This is a set (group) of conditions that must ALL be met correctly (i.e., must all return true) before the group can be said to be true. There could be zero, one, two,...infinite number of conditions within an AND group

  • OR Group also referred to as OR Connector

    This is a set (group) that combines two or more AND Groups of conditions. OR Groups needs AT LEAST ONE AND group in the set to be met correctly (i.e., one AND group to return true) before the group can be said to be true. There could be zero, one, two,...infinite number of AND groups within an OR Group

nlu details

Sample filter conditions

Sample 1nlu details
Sample 2 (NB: Selections could be made from available filter fields as well as filter values)nlu details

Test message response

Test that the set message responds when the right intent is picked.

nlu details

Question Response

Question Flow Tasks or Responses are used to give the user an opportunity to make selections from a list of already set options. These options each have a value that the ChatBot takes as a user input. The ChatBot either matches the input with an intent or as an entity keyword for the current intent.

nlu details

Test question response

Test that the set question is asked when the right intent is picked.
Important notes:
1. You should set all the filter conditions that ought to be met before submiting the filter. For the used example, no filter condition was set so the question is asked when the filter sequence is reached.


2. Flow tasks or responses are executed sequentially in the order that they appear on the dashboard. For instance, the 'Message Task' appears before the 'Question Task', hence, the ChatBot would first attempt running the Message if all the message's filter are met, then, it goes to the 'Question Task'. It goes like this for as many responses or tasks defined for each intent.
3. If a response from a user matches another Intent, the ChatBot would move onto the flows defined for the new Intent that was indentified in the user's input.

nlu details

Choose URL Response

URL options are a type of Flow Task or Response that allow users to specify multiple url links as options. It works pretty the same way as Question Responses but instead of matching options with new intents or entities, the selected url option redirects a user to a new website (as defined by the ChatBot owner).
A good example is when the ChatBot wants the user to read more information about a service available on some other website.

Select "Choose URL Option"nlu details
Specify URL optionsnlu details
Specify filter conditionsnlu details

Test URL response

Test that the set url option is accessed when filter condition is met.
Important notes:
A selected url option is opened in a new tab

nlu details

API Services

Application Programming Interfaces (APIs) allows one software application to communicate with another software application by sending and receiving information/resources through links called API endpoints.
API Services is a Flow Task that either sends (POST) inforamtion from your ChatBot to another endpoint or receive (GET) information from an endpoint to your ChatBot. Other operations such as 'Update' and 'Delete' can also be achieved using the right HTTP request between the ChatBot and other API endpoints (i.e., PUT, DELETE, PATCH, etc...).

Select "Make an HTTP Request"nlu details
Specify endpoint and data to be sent (i.e., payload)
The payload should be in the structure that it ought to be. If arrays, json arrays, strings, etc are the right payload, structure them appropriately before saving. nlu details

Test API response

Test that the set url option is accessed when filter condition is met.
Important notes:
A selected url option is opened in a new tab

nlu details

ChatBot Settings

General Settings

The ChatBot can be customized to have its own welcome text, theme colour, header text (Instead of "Chat") and Channels

The number of platforms available to the ChatBot depends on the subscription plan that was paid for.

  • Welcome Text

    This is the first message that a user gets from the ChatBot when the ChatBot is opened.

  • Theme color

    This is the colour your ChatBot apply to user's messages and web chat environment

  • Header Text

    This is the text used as indicator on web platforms to guide users on where the ChatBot is positioned. It is usually on the bottom right corner of the website.

  • Channels

    The Platforms through which customers can interact with it (i.e., Facebook Messenger, Twitter, WhatsApp, Skype, Web, Telegram, etc.). For a particular channel to become active, it's settings need to be configured through the "Channels" tab below the General Settings tab

General Settings
chatbot details

Channels Settings

Each Channel should be configured with appropriate credentials from the platform (i.e., Facebook, WhatsApp, Skype, etc credentials

  • Web

    Copy and paste the following script at the bottom of your website pages just before the closing body tag


    web widget snippet
  • FACEBOOK

    Create a Facebook App by visiting Facebook Developers Page

    Create new app
    chatbot details
    Enter App Name and Email
    chatbot details
    Copy App ID and paste on GladePay
    chatbot details
    Do the following sequence
    chatbot details
    Next
    chatbot details
    Next
    chatbot details
    Next
    chatbot details
    Next
    chatbot details
    Next
    chatbot details
    Next
    chatbot details
    Next
    chatbot details
    Next
    chatbot details

    GO TO SELECTED FACEBOOK PAGE, OPEN MESSENGER AND CHAT

  • Create a Telegram Bot by searching for @botfather in your telegram App

    Type the following commands
    • /help
    • /newbot

    Follow the instructions on Telegram to set up you new bot. Copy and paste bot access token on your Telegram Channel Settings


    Graphical Illustrations chatbot details
    Next
    chatbot details
    Next
    chatbot details
    Next
    chatbot details
    Next
    chatbot details