Genesys Cloud - Developer Community!

Hey everyone,

I don't quite understand why the Web Messaging send typing API requires a user context.

In our case, we're making internal API calls in the digital bot flow : we send the user input to a LLM, and wait for its response before replying. This call can take 5–10 seconds, and I'd like to show a typing indicator during that time, to let the customer know their request is being processed.

But:

• The API doesn't allow sending a typing event on its own (without a user context).

• There doesn't seem to be another supported command using sdk to simulate typing behavior.

• The only workaround I can think of is going through the Bot Connector, but that's not the architecture we've chosen.

Interestingly, when testing with AI Guide, it does show the typing indicator while the virtual agent is responding

Has anyone faced this? Any workaround or undocumented trick you've found to simulate typing from the server side? @Chad Hansen maybe ?

Thanks!

This Idea seems to be what you are wanting (though without requiring an API call) 

Hey everyone,

I don't quite understand why the Web Messaging send typing API requires a user context.

In our case, we're making internal API calls in the digital bot flow : we send the user input to a LLM, and wait for its response before replying. This call can take 5–10 seconds, and I'd like to show a typing indicator during that time, to let the customer know their request is being processed.

But:

• The API doesn't allow sending a typing event on its own (without a user context).

• There doesn't seem to be another supported command using sdk to simulate typing behavior.

• The only workaround I can think of is going through the Bot Connector, but that's not the architecture we've chosen.

Interestingly, when testing with AI Guide, it does show the typing indicator while the virtual agent is responding

Has anyone faced this? Any workaround or undocumented trick you've found to simulate typing from the server side? @Chad Hansen maybe ?

Thanks!

That's exactly it. I'll vote. Thanks for the answer!

This Idea seems to be what you are wanting (though without requiring an API call) 

Hey everyone,

I don't quite understand why the Web Messaging send typing API requires a user context.

In our case, we're making internal API calls in the digital bot flow : we send the user input to a LLM, and wait for its response before replying. This call can take 5–10 seconds, and I'd like to show a typing indicator during that time, to let the customer know their request is being processed.

But:

• The API doesn't allow sending a typing event on its own (without a user context).

• There doesn't seem to be another supported command using sdk to simulate typing behavior.

• The only workaround I can think of is going through the Bot Connector, but that's not the architecture we've chosen.

Interestingly, when testing with AI Guide, it does show the typing indicator while the virtual agent is responding

Has anyone faced this? Any workaround or undocumented trick you've found to simulate typing from the server side? @Chad Hansen maybe ?

Thanks!