Any suggestions on which integration to go with? My setup is as follows:
I have read over both Echosistant and Ask Alexa but can’t seem to understand which would be better for me.
Thanks in advance
Try them both. Play around. Both are being actively developed, so you will likely see improvements as Alexa improves.
I find that practical, everyday use is the best way to see which tools I need/use, and which are relatively useless due to functionality or convenience.
They are not mutually exclusive. Echosistant was born as an addition to more mature Ask Alexa app. Features within Echosistant are not available in Ask Alexa, although they can be achieved if one likes to tinker with the app with one major exception: the text free. Echosistant is capable of transmitting a full message to ST, which can be used for sms or to be broadcasted to Sonos speakers via text to speach service. Another feature of Echosistant that is not available in Ask Alexa, is the ‘Alexa Feelings’. You can use phrases like I am cold/ I am hot, is too bright, is not dark enough, that can easily be fine tuned in the app to suit your needs…
And at the same took me askAlexa has things you can’t find in EchoSistant. While on the surface they may seem like duplicates, they are not.
Both have their pluses.
And like @MEarly said, try both. I use both and I’m the
co-developer EchoSistant! !
Our goal is to enhance the ST experience… There are some exciting things in work.
Will having both installed and running at the same time not cause conflict?
Thanks for all the replies
The only conflict is that you cannot use the same invocation word. My Ask Alexa is SmartThings, while Echosistant is home, living room, kitchen, daddy, mommy and the list of invocation words with Echosistant goes on to cover each room in my home and each person…
The biggest problem is when I ask Smartthings (Ask Alexa) to turn the lights off in the Kitchen (Echosistant Profile). That’s when I blame the ST server for not working… 
I have askAlexa but not Echosistant. What does Echosistant have that askAlexa does not (or visa versa)?
The main thing that is that EchoSistant had free speech messaging. Which means you can say anything and that is what is sent. You do NOT have to have pre made messages.
Apologies for being a bit slow - but I don’t really understand the difference. If I tell you what I am looking for maybe you can let me know which one to go for.
I’m looking for natural interactions with Alexa. What I mean by this is none of the “Alexa, tell ‘name of skill’ to do this”. I’m looking for it to work like the native integration: Alexa, turn on lounge lights - however, with far more functionality.
From reading the above I though that EchoSistant appeared to be the closest to this - but then I started installing it and when I got to the lambarda script part the script was added as an “Alexa Skill”, rather than an “Alexa Smart Home” which made me think it would still rely on the “ask skill … or tell skill to do whatever”
Any clarification would be fantastic!
Sorry again
Each of the skills will require a “Alexa, ask SmartThings” or “Alexa, as Home” or some other invocation…you can’t get around this expect with the native Alexa integration.
not even by using the smart home sdk instead?
Yes, the SDK will allow you to access things directly, but only within the realms of what Amazon allows…so, on/off of switches…you would not be able to access things like locks or doors and use natural language (i.e. lock/unlock, open/close). If addition, you could not query the state of a device without using the on/off terminology…in which case you now have the equivalent of Alexa Helper, another app I developed before Ask Alexa…it allows you to control a variety of things, but again, you still need to use those in the context of on/off (i.e. Alexa, turn on night mode…this could run a macro called night mode that runs a variety of things, including locks and doors).
Make sense?
What EchoSistant does is allows you to build a room. For example, the living room. If you are sending a message you will say, Alexa, tell the living room… This is the message…
In the next version we have expanded what we call Alexa feelings and alexa thinks.
You can say things like, Alexa, tell home the kitchen is to dark. And the lights will turn on.
We are working on more personable interactions, expanding that and increasing reliability.
Unfortunately you still have to invoke the skill. But with EchoSistant you are able now, and even more so very soon, able to make it seem as though it’s a more direct integration.
Would you like to say, Alexa, tell the living room it’s movie time… AV changes to desired settings and lights adjust accordingly.
My favorite, Alexa, tell home it is too hot downstairs. And she adjust the thermostats accordingly.
We all struggle with this. I really don’t like telling “walls” to do something and more so my family members had hard time saying Alexa do something, let alone Alexa tell something to do something. Now like @MichaelS said, there are many advantages in using custom skills. Mainly that you can do things that native Alexa cannot.
With that being said, Jason and I are suckers for natural integrations. We want Alexa to work for us and not us to learn how to speak to Alexa. The closest we came to more natural language was to create a bunch of skills tied to the same Lambda code and invoke them as we need.
I haven’t really explored the Smart Home Skill API because I am deeply involved in making the custom skills work for me AND that so-called skill adapters in the Smart Home Skills don’t seem all that flexible.
I can tell you though, that based on Amazon documentation, you should be able to call a skill without using connectors like “tell, use or ask”. However, I didn’t have much success in my tests nor had time to research what it takes to call a custom skill without those connectors (more to come on this).
Thanks for all the responses guys. Looks like this is an Amazon issue as it’s not possible yet.
You have all made some amazing extras with being able to query device states and infer from feelings… but unfortunately the invocation word is not going to pass the wife test for me even though I do want all those extra features.
Currently I just create virtual on off devices and then either set a core action or a custom smartphone.script off them being flicked on / off so that I can get extra functions from. Alexa. So I can say things like “Alexa, turn on the cinema” and a custom script I wrote will check the state of each of the cinema items (all individual speakers, av receiver, boxes, projector, projector screen, projector lift etc) and then turn on, in sequence with correct gaps all the components that are needed). For the times where saying “turn on/off” doesn’t make sense I use ifttt so that I can say "trigger the"
it’s working well so I’ll wait for Amazon to make the smartphone SDK have more options before I change approach.
What more natrural than saying:
Alexa turn the lights off in my home
Alexa turn the light on/off in the bedroom
Alexa tell me about the garage door in my home
Alexa what is the temperature in my home/bedroom/living room?
Alexa what devices are inactive in my home?
Alexa what devices have low batteries in my home?
Alexa what lights are on in my home?
Alexa turn the TV on/off in the x room?
And how about restricting guests to control your home via Alexa?
Alexa enable the pin number for thermostats, locks, garage doors, etc… in my home!
Keep an eye on EchoSistant v4, a custom skill that breaks the norm!
sounds great
@SBDOBRESCU just to clarify, each of bedroom, living room, x Room you mention above would all have their own profile? Is that in the current EchoSistant or v4? I think I just got my eureka moment for understanding profiles
Thanks
Yes, each of those are a profile. EchoSistant was designed around Profiles. We are just taking them to the next level with v4…
Question for those of you that were debating which route to go…
Which did you decide to go with?