📞Voice assistant

From start to finish, set up a voice-based bot in minutes with NLX's Dialog Studio

Checklist

You'll complete the following to successfully launch your voice assistant:

Pre-setup: Integrations

Est. time to complete: ~8 minutes

A one-time integration of a Natural Language Processing (NLP) engine must be completed in your workspace:

A one-time integration of a voice-enabled communication channel must be completed in your workspace:

Step 1: Create an intent

Est. time to complete: ~4 minutes

Begin by identifying the tasks your AI assistant will automate and organize them into individual topics called intents.

Each intent is invoked when your chosen AI model matches a user's query ("I want to order room service") to an intent you've created (OrderRoomService). This match is done based on the training data you provide for your AI model.

  • Select Intents from workspace menu > Choose Create or Add intent

  • Enter a descriptive name (no spaces or special characters) > Select Save

  • Click + Add new training phrase > Create 5 phrases that users might say to invoke the intent

  • Select Attached slots > Attach custom and/or built-in slots that will be resolved in your workflow

  • Click Save

  • Select + Add flow > Click Create flow (ensure your chosen voice channel is set as the Primary)

  • Click Save

Repeat for any additional intents your bot will help automate.

Step 2: Construct a workflow

Est. time to complete: ~15 minutes

Determine the sequence of steps and messaging that a bot must follow to assist a user with the task. The conversation workflow is assembled in an intent's canvas builder with a pattern of nodes similar to a flow diagram:

  • Select the flow card created in Step 1

  • Add nodes by right-clicking on the canvas & choosing New node (see available node types)

  • Connect first node to Start node > Connect sequential nodes via node edges or stacking

  • Add bot messages to nodes by selecting + Add message on a node's side panel

  • Click Save

Step 3: Set up bot

Est. time to complete: ~3 minutes

Now you'll get to create the AI assistant that users will interface with. This step involves attaching all intent workflows you want your bot to access, defining flows to handle certain behaviors, and setting up the channel your bot supports.

  • Select Bots from workspace menu > Choose Create or New bot

  • Enter a descriptive name > Click Save

  • Click Intents tab of bot > Select + Add intents > Attach one or more intents you've created to make available to your bot

  • Select Default behaviors tab of bot > assign any attached intents to the bot's behaviors > Click Save

  • Select Channels tab of bot > Expand the voice channel your bot will support (e.g., Amazon Connect, Amazon Chime SDK, etc.). Be sure this is the same channel set on your intent flow(s) attached to the bot, or add the channel to your intent flows > Click + Create channel

  • Enter required details for voice > Click Create channel

Step 4: Build & deploy

Est. time to complete: ~5 minutes

A build constructs the array of workflows that make up your bot and updates any changes made to your flows, while deploying makes a successful build live:

  • Click Deployment tab of bot > select Create or Review & build

  • Wait for validation to complete > Select Create build*

  • When satisfied with a successful build, click Deploy

Optional: Retrieve data

To pull in and reference information from an external system, you may first set up a Data request integration and invoke it in a workflow using the Data request node.

  • Select Data requests in workspace menu > Choose + New data request > Provide a name

  • Switch to External implementation under the Implementation tab > Choose the appropriate HTTPS method

  • Expand Production endpoint > Enter the URL and any applicable Header(s)*

  • Download the provided AWS CloudFormation template listed in the Instructions section

  • Select the Response model tab > Choose <> Auto-generate option > Enter a snippet of your schema

Optional: If you require sending a payload when the API call is made in a workflow, select the Request model tab > Click Create from example JSON > Enter a snippet of the payload schema

  • Click Save

Now that your integration is set up, trigger your API call during an intent flow using a Data request node

Last updated