How to Make an AI Chatbot Without Coding and Deploy on Websites and Apps
How to Make an AI Chatbot Without Coding and Deploy on Websites and Apps
In this post, you will learn how to make a chatbot for your website and apps using Kommunicate's simple and NLP-powered chatbot builder - Kompose.
Join the DZone community and get the full member experience.Join For Free
In this guide on how to make a chatbot, we will be learning how to make a chatbot using Kompose bot builder. You will learn:
- Why are chatbots important? More so in 2020
- How to create welcome messages and intents
- How to use rich media such as buttons, lists, and card carousels
- Triggering chatbot to human hand-off
- How to make multilingual chatbots
Why Do You Need a Chatbot?
With the increasing digital-savvy customers and businesses across the world. There’s a need for the chatbot to drive customer communication, enhance customer loyalty, and improve business growth. Also, chatbots automate and reduce the number of tickets/conversations received on the customer support helpdesk and the workload on humans.
How to Make a Chatbot
Conventionally the chatbot can be built with chatbot building platforms such as Dialogflow and AWS Lex. These platforms provide natural and rich conversational experiences. But it really needs a good understanding and knowledge of the building platforms to create a good chatbot.
We have guides for you to look into if you want to explore the platforms mentioned above:
On the other hand, Kommunicate’s bot builder - Kompose is a solution to all the people - a techie, non-techie, or a person who doesn’t have any idea about chatbots. Anyone can create a chatbot with the Kompose without any assistance.
How to Create a Chatbot Using Kompose Bot Builder
Kompose is a GUI bot builder based on natural language conversations for Human-Computer interaction. You don't need any coding skills to master Kompose. It has a simple, intuitive, and easy-to-use interface.
In this post, we will learn how to create a chatbot using Kompose quickly.
Bot Integration With Kompose
Navigate to the 'Bot Integration' section >> Select Kompose >> Setup Bot
Sections in Kompose
There are four primary sections in Kompose. Let me introduce you to them, and then we will dig deeper for each of them.
- Welcome Message - This is the first message your users will see in the chat widget
- Answers - This is your primary workspace. Add the user’s message and bot replies here. You can make all your conversational workflows here
- Small Talk - It’s an in-built set of general replies & chit-chat
- Fallback - The fallback message when the chatbot is unable to understand the user’s queries
The welcome message is the first message that the chatbot sends to the user who initiates a chat. The welcome messages create the first impression, but they also maintain customer engagement and make them feel valued. An efficient bot greeting message also decides the conversational flow with visitors and giving a chance to explore the company’s products & features.
You can set an interactive greeting message using the Kompose builder with the help of the rich messages - Buttons, Images, or any other rich media type. An example of creating a welcome message in Kompose with the buttons:
Answers section is where you've to add all the user's messages and bot responses. It is similar to the intents section in the Dialogflow.
In the ‘Configure user’s message' section - you need to mention the phrases that you expect from the users that will trigger. Note that, Kompose runs on powerful NLP, so the bot will be able to match intents for similar phrases as well. But as you know - more the data - better the accuracy. So make sure you add more phrases to fine-tune the bot.
In the ‘Configure bot’s reply' section - you need to mention the responses (Text or as Rich message) the chatbot will deliver to the users for the particular user’s message. You can add any number of answers and follow-up responses for the chatbot.
In the above welcome message example, we’ve added buttons 'Book a Demo' and 'Give me a product tour.' So for these two buttons, we have added answers on the bot reply section.
On clicking that button, it will show that particular answer as a response to the user. In this way, you can add all your text and rich messages as answers and follow-ups. You can check your intent on the fly by clicking the 'Test' button.
It’s an in-built set of general replies. Small talk improves the end-user experience and to match the authentic human interaction. We all know that people play around with chatbot at times. Hence, these small talk messages make your chatbot foolproof by accomodating general questions. You can edit them and also add custom small talk messages by clicking the 'Add+' button and mentioning the user messages and their corresponding bot reply.
Fallback/Unknown User Input
You can also include Fallback Responses (bot reply) in the event the chatbot comes across an unrecognized user's message.
In the below example, the bot’s fallback response is set with the image. Likewise, you can set it with just text, buttons, or other rich messages. When an unrecognized user's message 'fallback' is entered, the bot forwarded the conversation to a human with the 'Fallback message.'
How to Use Rich Messages
Rich messages provide users with a smooth, convenient, and quick way to interact with your business products and features. Rich media increase your click-through rate, and it’s user friendly.
Kompose supports the following rich messages - Buttons, Images, Lists, Cards, and Forms. Please refer to this document to know more about the rich messages and get the custom payload JSON examples.
For creating rich messages in Kompose, under the “Answers” section, click the ‘Add+’ button. Then configure the user’s keyword message and then click the button or images or custom payload [For Lists, Cards, Forms], which will be under the “Configure bot's reply” section and then Save Changes.
For buttons, just type the name of the buttons you want, and you can create as many buttons as you wish.
In the Image URL section, use the image URL the chatbot should show.
Here is how the images will come up in your chat widget.
Custom payload lets you create a plethora of rich message types. Paste the custom payload JSON code in the ‘Custom’ - ‘Configure bot's reply’ section. Please refer to this doc to get the custom payload JSON examples.
Here is how the above carousel will render in your chat widget:
Transfer to a Human Agent
The chatbot to human hand-off plays a significant role in customer support and critical situations. It is useful when your chatbot is unable to understand the user's query and fails to answer them.
The bot to human handoff should occur on tricky & tough situations, keyword triggers, and based on user requirements.
In Kommunicate, there is an option for the bot to human handoff, Manage Bots >> Edit >> Human Handoff.
If you enable this, the chatbot will automatically hand-off conversations to humans in case it is not able to answer them. The fallback message described in the above section will be triggered.
In Kompose, you can also configure a 'Talk to Human' button with custom payload and transfer it to a human agent.
For example, when the user clicked on the button "Talk to Human," the bot transferred the conversation to an agent.
How to Make a Multilingual Chatbot
A multilingual chatbot is much needed for a business to pull audiences around the world. If a chatbot doesn’t support multiple languages, then it restricts your business boundaries. Kompose chatbot supports numerous languages.
In your Kompose bot, under the “Answers” section, click the ‘Add+’ button. Then configure the user’s keyword message and the corresponding bot's reply for the message in your language. In the below example, Spanish is used, and the Kompose bot correctly responds to the corresponding user message.
To sum it up, it is pretty easy to make a chatbot using Kompose bot builder. Go ahead and play around with Kompose and create your workflows and chatbots easily. In case you want a video tutorial, here is one:
Published at DZone with permission of Parth Shrivastava . See the original article here.
Opinions expressed by DZone contributors are their own.