Genesys Cloud - Main

Hey team,

We have implemented Speech Recognition on our main IVR.
We have been issues where the customers arent able to utilize the IVR in a convenient way.

Our biggest issue with Speech Recognition:

• Customers are getting routed to our default queues whenever there is a bit of a background noise.
• Sometimes even when they cough or talk in the background, its getting routed to default queues and not letting customers choose any option.

Its creating a very unpleasant experience for our customers.
Our current Minimum confidence level is set at 40%.

Has any other partner implemented this feature and found any ideal %.

Thanks,

First, I would say that calling the speech menus speech recognition is like calling this aTesla

I would suggest you completely remove all that junk and create a simple bot to do that same thing.  It would take very little effort and create an entirely different customer experience.  That menu-based recognition is nothing more the ININ Speech in the old CIC, which is nothing more than sound matching or what is called phoneme isolation where the sound that words are made up are matched against what is heard, including any background noise.  To say it stinks would be very optimistic.  If you need to know how to create a simple NLU menu using a Genesys Dialog Engine bot.  

Hey team,

We have implemented Speech Recognition on our main IVR.
We have been issues where the customers arent able to utilize the IVR in a convenient way.

Our biggest issue with Speech Recognition:

• Customers are getting routed to our default queues whenever there is a bit of a background noise.
• Sometimes even when they cough or talk in the background, its getting routed to default queues and not letting customers choose any option.

Its creating a very unpleasant experience for our customers.
Our current Minimum confidence level is set at 40%.

Has any other partner implemented this feature and found any ideal %.

Thanks,

First, I would say that calling the speech menus speech recognition is like calling this aTesla

I would suggest you completely remove all that junk and create a simple bot to do that same thing.  It would take very little effort and create an entirely different customer experience.  That menu-based recognition is nothing more the ININ Speech in the old CIC, which is nothing more than sound matching or what is called phoneme isolation where the sound that words are made up are matched against what is heard, including any background noise.  To say it stinks would be very optimistic.  If you need to know how to create a simple NLU menu using a Genesys Dialog Engine bot.  

Hey team,

We have implemented Speech Recognition on our main IVR.
We have been issues where the customers arent able to utilize the IVR in a convenient way.

Our biggest issue with Speech Recognition:

• Customers are getting routed to our default queues whenever there is a bit of a background noise.
• Sometimes even when they cough or talk in the background, its getting routed to default queues and not letting customers choose any option.

Its creating a very unpleasant experience for our customers.
Our current Minimum confidence level is set at 40%.

Has any other partner implemented this feature and found any ideal %.

Thanks,