How to create a block in a chatbot to detect customer's intent
How do I create a block in a chatbot to detect customer's intent?
This article explains how to set up a chatbot to analyze user messages to find out what they want. Follow these steps to train your chatbot with example messages. This helps the assistant match customer phrases to correct outcomes.
In this article, you will learn how to
- Add a Detect intent block to your chatbot.
- Provide example messages for customer phrases.
- Fix intent matching errors.
Before you start
Before creating a block in a chatbot to detect customer's intent, make sure you have an existing chatbot to add a Detect intent block.
How to create a block in a chatbot to detect customer's intent
- Log in to your Flownally workspace.
- Go to the Chatbots tab.
- Click an existing chatbot to edit it, or create a new one to open the editor automatically.
- Click on Add step after a block, or on the + icon on a connecting line, to add a new block.
- Choose the Detect intent block.
- In Message to check, choose if you want the chatbot to Wait for a new message or Use last message.
- In Intents, add the Intent name.
- In Example messages, provide examples of messages you might receive from a customer that the chatbot would use to recognize this intent.
- Optionally, click Add intent to add more intents and split your chatbot into more paths.
- Add the next steps for each intent on the canvas. Messages that match no intent follow the No matching intent path.
What happens after creating a block in a chatbot to detect customer's intent?
After creating a block in a chatbot to detect customer's intent, your assistant can interpret text messages. The system uses AI to process customer inputs and routes them to the path of the best matching intent.
Troubleshooting
Wrong intent triggered
This can happen because your example messages overlap too much. To fix this you need to edit the phrases to make them distinct.
Chatbot can't be activated yet when trying to activate the chatbot
This happens because an intent has no name, for example Intent 1: Name is required. To solve this problem, you should provide a name for each intent. Also add examples of what the customer might say, so the chatbot recognizes the intent.