Let us use a programming language to control the conversation flow instead of forcing us to use the dialog editor

The main feature of Watson Assistant is to categorize utterances and extract entities from them. Currently, the only way to use this feature is to edit the dialog using the UI provided on IBM cloud. It would be nice to have access to the main feature by having an API call that only returns the intents and entities of an utterance. The main reason why this would be useful is to implement chatbots using programming languages such as node.js or Python. The advantage of this would be to make the organization of the conversation simpler than by using nodes and make more complicated conversations simpler to program.

For example, think about how you would implement the following using nodes:

WA: "What items do you want?"

Customer: "I want 3 fries, a coca-cola and 2 cheese burgers"

WA: "I see you ordered 3 fries, 1 coca-cola and 2 cheese burgers"

Using the dialog editor, this is almost impossible and would require a douzain nodes to implement the logic and it would make the dialog super messy. It also defeats the purpose of thinking about nodes as 1 condition 1 response. Using a programming language, it would be much easier to see that "@sys-number:3" comes right before the entity "@menu-item:fries", the entity "@menu-item:coca-cola" doesn't have a number before it so it must be qty 1, etc. This is just one instance where a programming language would be useful but there are many more.

I tried to implement a solution that does that myself using the current version of the API but I feel like I'm swimming against the current. Can you please add API calls that would make it easier to do design a dialog using a programming language?

Let me know if you want more information

Thank you