I went this way after the Heroku server changes in Echo Speaks a while back. I really don't miss the futzing around with Echo Speaks.
For my 3-4 things I want spoken, it was quite easy to just create an Alexa routine, and form a couple rules in Hubitat to trigger those routines. Like HAL says, I would suggest using this method over E.S. if you only have a few basic items that need announcements, it is a much simpler method to achieve the same goal.
The community driver Universal Virtual Device Switch uDTH helps tremendously to change one device type to another, you can use this virtual device to easily create a contact sensor for Alexa.
The downside is that this all depends upon Alexa working properly, and there have been a handful of times over the past couple years where that routine did not function properly, apparently due to issues on Alexa's side of things. But I am not sure Echo Speaks would have been any different if the error was on Alexa? Either way, none of the stuff I am announcing is for critical life/safety stuff, and I also have text messages as a backup method.
