January 30, 2020

The three reasons to build your Dialogflow chatbot with Ideta.

What is Dialogflow ?

Dialogflow is a powerful tool developed by Google. It does Natural Language Processing. This is a field of artificial intelligence that enables machine to understand human language. Developers use Dialogflow in order to be able to extract information from user input.

For example, if I am a retailer and one of my customers says "I want to buy an apple", I actually want to add an apple to the customer's basket. So I need an action and some details about this action. Basically, I need to know that :

  • I must add something into the customer's basket -> that's the intent
  • What something to add into the basket ? -> that's the entity

That's what Dialogflow does ! It extracts information from text and changes it into something that a website or another computer system can use.

By using Dialogflow, developers do not have to build their own machine learning models. It is basically a tool that enables developers to do machine learning even if they don't have the time or the ability to do it themselves.

What Dialogflow does for developers, Ideta does it for you. And this is what you gain from it :

  • lower development costs
  • easier project management with foreseeable deadlines and outputs
  • lower maintenance costs
Ideta helps you leverage artificial intelligence via Dialogflow without coding.

In detail, this is how it works.

Step 1 : You enter training sentences similar to the one you would like your chatbot to understand, just like the image below. Dialogflow will understand more than this set of sentences, and if he thinks the user input is similar, it will send the intent it detected back to us. So, he the user says "how much does it cost ?", he will answer something meaning "intent price detected".

Set of training sentences

Step 2 : You hover over part of the sentences that are interesting. Dialogflow will extract similar stuff. So if the user says "My name is Yanis", he will answer something meaning "intent first name detected. Value of first name is Yanis". Replace Yanis with Sarah and you will have the same intent detected but a different entity.

Full names being extracted

Step 3 : Now your training is done. You can export it to Dialogflow. In the video, you can see there is a file to upload. This file is obtained following the following procedure -> https://dialogflow.com/docs/reference/v2-auth-setup

Step 4 : Then you just need to select the intent, the entity into a dropdown list and then you can answer your customer appropriately.

Detecting intents in Ideta

You can just as simply store the values extracted by Dialogflow and use store them for later use.

Storing NLP values for later use with Ideta

And that's it ! With these simple steps, you know everything there is to know about building a smart chatbot with Ideta and Dialogflow.

You can watch the video recap here :

You are now ready to automate your conversations and increase your customer's satisfaction rate. Sign up to start automating your chatbots with AI.

You can contact us if you still have questions.

You can as well register to our webinar if you want an interactive moment to ask your questions while watching a live tutorial -> It's here !