question

Matt Kruse avatar image
Matt Kruse asked

Can I Test Without Talking To My Echo? (launch requests via the web?)

My Echo is at home, but I'd like to do some testing from work. Is there any way to launch a request, perhaps from the echo.amazon.com site? Or to re-try a previous command from the list? It would be great to have some kind of online "tester" wouldn't require the echo to run, but would hit the endpoint and interpret the results just like the Echo does when being spoken to.
alexa skills kit
10 |5000

Up to 2 attachments (including images) can be used with a maximum of 512.0 KiB each and 1.0 MiB total.

Jiggybyte avatar image
Jiggybyte answered
A "virtual Echo" would be fantastic. Not only would it allow us to develop/test remotely but it would allow us to develop/test while someone else is using the echo in our house. Then my wife won't get mad when I tell her she can't stream her music tonight while I am tinkering with the development of my application. :-) An interface from the web that allows me to type in a string of text "to be interpreted" by the echo, and then passed through the site to our endpoint doesn't seem too complex on the surface of things, and would certainly speed initial development. Of course, at some point we need to actually test the vocal part to make sure the commands are getting heard/processed properly, but I do think this is a great idea.
10 |5000

Up to 2 attachments (including images) can be used with a maximum of 512.0 KiB each and 1.0 MiB total.

Nick Gardner avatar image
Nick Gardner answered
Hi, There's no interface to do it, but if you disable signature checking, you can just curl directly against your endpoint with the correct data and see what you're getting out of it. The one thing this doesn't do is the filtering that the actual Echo would do when trying to parse your voice. As always, thanks for your feedback. -Nick
10 |5000

Up to 2 attachments (including images) can be used with a maximum of 512.0 KiB each and 1.0 MiB total.

Jiggybyte avatar image
Jiggybyte answered
The more valuable piece of the virtual echo or whatever you would call an interface like this is parsing our grammar and intents and what it on your (amazon's) end and getting the Json from you as if through our Echo. I can write tests locally all day, but if something upstream is parsing wrong then I'm testing the wrong thing. Just my two cents, this would be very nice. Especially for people who don't yet have an echo but want to write an app.
10 |5000

Up to 2 attachments (including images) can be used with a maximum of 512.0 KiB each and 1.0 MiB total.

The Stig avatar image
The Stig answered
Here you go Matt, Another developer coded this - https://github.com/jjaquinta/EchoSim/releases It's a virtual Echo that lets you type in stuff, then it responds as if you had said actually said it.
10 |5000

Up to 2 attachments (including images) can be used with a maximum of 512.0 KiB each and 1.0 MiB total.

jjaquinta avatar image
jjaquinta answered
It needs a lot more features! But I'll pivot back to that once I get an actual Echo service deployed. AWS certificates were giving me headaches today. Feel free to enter in bugs and feature requests to the issue tracker!
10 |5000

Up to 2 attachments (including images) can be used with a maximum of 512.0 KiB each and 1.0 MiB total.

Amazon Customer avatar image
Amazon Customer answered
I don't know why Amazon cannot provide a simple web app or mobile app that can be run from your desktop/laptop or phone which already has a mic and we can see what JSON the Echo app would output based on your speech. Most of the threads I have seen on this topic is about to test out web end point. There are 100s of way to do that. The answers are missing the point. The new thing here is how Echo translates your voice to a request to your webservice. This is the most critical test environment needed. Can't Amazon developers right one soon and publish it like tomorrow? Or is this a more strategic decision of someone in product marketing to sell more Echos and not let other developers integrate the Amazon Speech APIs in their applications? If you have a lot of friction for developers to develop Echo Apps, then you are not likely to see the number and quality of Apps you expects from the developer echo system. Or worse yet someone from another company will make this happen and create another portal that's better for developers. Please give this a priority and convince the people in management to publish this soon.
10 |5000

Up to 2 attachments (including images) can be used with a maximum of 512.0 KiB each and 1.0 MiB total.

jjaquinta avatar image
jjaquinta answered
It's called Amazon Voice Service. :-) Seriously, though, the STT Echo uses is quite complex and proprietary. Right now it only works on one device. Clearly it's non-trivial to get it to work on more than one, otherwise they wouldn't be making a big effort with AVS. That is, almost certainly, going to be marketing's answer: "Wait for AVS, then do whatever integration you want with whatever device".
10 |5000

Up to 2 attachments (including images) can be used with a maximum of 512.0 KiB each and 1.0 MiB total.