Not able to type any keyword when pick an option is visible in Virtual Agent

[pooja_m]

Hi All,

I have created a custom topic named as "Main Menu". When I select Main Menu, it shows an options to pick. at same time, typing a keyword gets disabled until unless I pick any option.

I want to give both functionalities to customer(either pick an option or type it manually at same time). Can anyone please help me out, if there is any way to achieve this?

Thank you in advance!

[Chris D]

Yes, you can achieve this by not doing a custom topic at all.

Presumably, your three options are three different topics. If they're not, you should very strongly consider it because that will allow you to use the ootb topic picker ("Show Me Everything") and lets the user also type in what they want.

You have fine and very easy control over which topics appear in the topic picker so even if you have more than three topics but you want to funnel all users through one of these three, you can hide all topics except for these three.

[pooja_m]

Yes, three options are three different topics. if I pick any one of the option, respective flow gets triggered. In that flow i have created one static choice as "select quantity" with options as "1. one", "2. many". In that case,  i am not able to type a keyword "many" manually instead of selecting an option (many) from the drop down.

I wanted to achieve typing a keyword when static choices are shown in bot. is it possible in VA?

Thank you!

[Chris D]

As Lynda said in a different response, you can try to select the option "Enable NLU at Input Node":

Users can enter text to select items or switch topics from this node when this switch is enabled.

I didn't put this option in my initial response because I tested it out and this use case wasn't working for me. Plus, even if it was working, I don't like the idea of enabling NLU at random inputs - it can easily cause unintended consequences (switching topics), especially if your model has a relatively low confidence threshold and/or triggers topics with one word utterances.

I suppose there's a low risk if you're using it for Yes/No inputs ("yes" and "no" certainly should not trigger any intents) or other very simple options ("one" vs "many" seems safe as well), so you can try it out yourself and see if it works for you.

[Lynda1]

Chris D. suggestion is the best.  if you are instant on this approach, you can turn "Enable NLU at Input Noe"