Everything you need to manage a chatbot!

The dydu solution allows you to manage all the stages of your chatbot's life cycle. With more than 10 years of experience with large and demanding customers, the functional richness of our solution makes it possible to cover all needs. No technical expertise required.

Test the solution

Dialog engine

The dialog engine and natural language processing are essential to bring your chatbot to life. Without them, there is no conversation and therefore no chatbot.

Proprietary natural language processing

Since the beginning, dydu has had its own linguistic brick, we aren’t based on an external actor for natural language processing. This allows us to be completely independent. Our linguistic brick manages French particularly well and is able to correctly manage ten languages.

In a simplified way, our language brick calculates a distance between the user's request and all the formulations included in your chatbot's knowledge base. It is only when this distance is small enough that the chatbot is able to provide an answer to a question. This makes it possible to ensure a much greater relevance of responses than a system based on keywords. This mode of operation also makes it possible to quickly identify queries that still need to enrich the knowledge base without having to constantly check what is already working.

Decision trees

Decision trees

To provide answers to user questions, chatbots need content, so it is important to be able to feed this knowledge base easily. We therefore propose a simple tool allowing to design basic knowledge and decision trees. The branches of these trees can depend on a complementary response from the interlocutor but also on the result of the call to an external web service or on information that the dialog engine would have concerning the interlocutor during the conversation.

Web services consumption

Web services consumption

Providing answers to questions is great, but allowing personalized answers is even better. With the consumption of web services from your chatbot's knowledge base, you can both consult external information such as for order tracking and push information outside to feed a CRM or a ticket management tool, for example.

The configuration of these web services is done directly from the dydu solution. Whether your web services produce XML, JSON or even plain text, whether they require authentication or not, it doesn't matter, you will be able to access them from our solution.

Integrations

Your chatbot must be available to your users where they need it, which is why we offer several integrations

Web

Web

Web integration allows you to provide your users with a dialog box on your website or on your intranet. Nothing could be simpler, just copy / paste a line of code to include the javascript of the dialog box in your pages.

You can customize the appearance of this dialog box to match your design guidelines.

This dialog box includes the following features, among others :

  • Onboarding process
  • Rich content
  • Side panel in addition to the text response
  • Retrieving user reviews

Instant messaging

Instant messaging

Your chatbot can also respond to users from different instant messengers, including :

  • Facebook Messenger
  • Microsoft Teams
  • The Microsoft Bot Framework which allows you to communicate with many other channels

Activity monitoring

Your chatbot will live, you must be able to follow its activity and evolve the content of its knowledge base in order to increase its rate of relevant responses as well as the satisfaction of end users.

Dialogs reading

Dialogs reading

Reading your chatbot's conversations is a tedious but necessary step. You don't have to read the entire conversation, it becomes impossible by force of circumstances if your chatbot is in high demand. You can apply different filters to focus on the conversations that interest you.

Users have the possibility of leaving an opinion at the end of each restitution of knowledge during a conversation, this makes it possible to feed the analytics and to identify the subjects generating dissatisfaction, but it also makes it possible to read only conversations containing negative opinions, for example.

You can directly from the conversations enrich the knowledge base of your chatbot, either by creating new knowledge but also by modifying existing knowledge.

Analytics

Analytics

Many analytics are natively present within the dydu solution. They allow you to follow the requests of your chatbot, whether in terms of volume or in terms of the subjects discussed.

You can thus know which knowledge or topics are the most recurring, those which generate the most satisfaction or dissatisfaction.

You can also follow the response rate of your chatbot and see how it evolves according to the enrichments you bring to your knowledge base.