If not for the budget, Could have nearly achieved a voice based provisioning platform - CWM Extension

Following would be key steps/aspects 

1) Open Standards in providing NLP inputs ( any RFC present ? yes Ref[1] ) 
2) Mapping the voice input to the back end business workflow use case
3) Identify additional input and prompt it with user who can pick it in interactive manner
4) Validate input in step 3) 
5) Construct JSON and invoke REST API
6) Invoke the workflow as in existing system today.

This would have meant a voice-ready remediation platform.
A feature that BMC remedy product would also need to think off in near future.

Anyways. Some one else will do it.

References:
[1]
https://tools.ietf.org/html/rfc4313

Good reading(s)
[2]
https://www.softwaretestinghelp.com/voice-recognition-software/
https://www.totalvoicetech.com/how-voice-recognition-technology-works/

[3]
Voice sense - Research
https://www.readingrockets.org/article/speech-recognition-learning

Remember the Voice Configuration Assitant hackthon - 2016
This could have been a feature in CUCM. One that I repent not doing it.

Also Microsoft O365 has voice assitant who subsitutes stenographer work of older days.
:)

[4]
https://www.twilio.com/bots

Transactional bots are like goldfish – they don’t remember previous interactions with the user and can’t maintain extended dialogue with the user.
 

Conversational bots are like elephants – they maintain the state of the conversation and carry information between turns of the conversation.

[5]
https://www.drift.com/wp-content/uploads/2018/01/2018-state-of-chatbots-report.pdf