Channels
SARAH is a SmartHome opensource project created in 2012 to control internet of Things with voice commands. It was build on top of Microsoft Speech Recognition Engine (and require Windows).
The windows release of VISEO Bot Maker provide a default Node-RED flow with SARAH nodes to start quickly. Follow the SARAH node's documentation.
The VISEO Bot Maker provide a wrapper of MS Bot Framework which handles up to 12 different channels including Facebook, Skype and Slack.
The MS Bot Framework will
- translate channel message to generic message
- forward messages to/from your server
That's why your server MUST be available online throught an HTTPs connection.
To get started, you first need to create an account for the MS Bot Framework (or sign in with your existing account).
Then create a new bot.
When prompted, choose "Register an existing bot built with Bot Builder SDK".
This MS Bot Framework app defines your bot. You can add as many channels as you want, MS Bot Framework will redirect all the messages to a unique messaging endpoint, to your bot server.
To define your "messaging endpoint", edit your app settings and enter for messaging endpoint
https://[your server url]/api/v1/messages
Finally, you need to register your bot with Microsoft and generate an App Id and password.
The MS Bot Framework won't store the password for you so you need to save it somewhere.
Now you must configure your server to receive messages.
In your server UI (https://[your server url]/), add the following nodes to your flow.
or import the following json encoded text by choosing Menu > Import > Clipboard.
[{"id": "61f5edf4.da6b4c", "type": "tab","label": "Flow 1"},{"id": "4d614ba5.4753cc","type": "bot","z": "61f5edf4.da6b4c","x": 200,"y": 100,"wires": [["c4b44d72.5df2e"]]},{"id": "c4b44d72.5df2e","type": "send-card","z": "61f5edf4.da6b4c","name": "","prompt": false,"promptText": "","speech": false,"speechText": "","delay": 0,"assert": "","sendType": "text","text": "Hello, I'm a bot !","random": false,"media": "","title": "","subtitle": "","attach": "","carousel": false,"buttons": [],"quicktext": "","quickreplies": [],"btnOutput": false,"quickOutput": false,"repeat": 0,"outputs": 1,"signintext": "","signintitle": "","signinurl": "","bt1action": "postBack","bt2action": "postBack","bt3action": "postBack","quickbt1action": "postBack","quickbt2action": "postBack","quickbt3action": "postBack","quickbt4action": "postBack","quickbt5action": "postBack","quickbt6action": "postBack","quickbt7action": "postBack","quickbt8action": "postBack","quickbt9action": "postBack","quickbt10action": "postBack","quickbt11action": "postBack","x": 580,"y": 100,"wires": [[]]}]
You now need to tell your bot which MS Bot Framework app it should respond to.
Double-click on the Bot Server node and enter the App Id and Password you previously saved from the MS Bot Framework.
Now your server is all set ! You can already test by going in your MS Bot Framework app and click on the "Test" button. This will open a webchat panel with which you can speak with your bot.
You can now add your channels in the "Channels" tab (Facebook Messenger for instance) and follow the procedure explained in the form.
Work in Progress
The VISEO Bot Maker provides an open API to connect custom channels.
- The Bot Wrapper can build a Message from input Data then handle response from
Messagenodes - The Dialog Flow (aka API.ai server) node is an implementation of the Bot Wrapper dedicated to Dialog Flow (and so Google Home)
- The WeChat node is an implementation of the Bot Wrapper dedicated to WeChat (and chinese market)
More implementation like Alexa are on the way. Implementations are performed aside customer projects.
The VISEO Bot Maker provides lots of nodes to manipulate and design conversation.
- Speech Recognition like Bing Speech, Google Speech, ... use AI to convert Audio Buffer to raw text
- Natural Language Processor (NLP/NLU) like DialogFlow, LUIS, ... use AI to convert raw text to intents and entities
- Computer Vision like MS Cognitive Services, Google Vision, ... use AI to convert images into meaning
- Analytics like FaceBook Analytics, Azure App Insight, ... forward bot behavior to dashboard to manage KPI.
VISEO adds and improves nodes for it's customers while designing their bots.
- the Salesforce node helps bots to retrieve data from salesforce to perform care level 1.
- the Sensit node monitor IoT device to perform proactive actions