You built your first action with watsonx Assistant in Part I, then refined your content in Part II. Now let’s learn how to test your assistant and deploy it on your website! 

Publish and preview your assistant’s content 

Go to the Task tracker, click the arrow next to Deploy your assistant to a test channel, then select Publish your latest content to the live environment 

Click the blue Publish button at the top of the table, and voila: you’ve published your first version of your assistant’s content! Your assistant is not visible to customers yet (we’ll cover that process soon), but this version is preserved, and any subsequent edits you make will be saved in the draft environment. 

Now it’s time to share this first version of your content with your colleagues! Click the Preview icon (the play icon) to open up the Preview page.  

The left-right navigation has three key links: 

  • Change background website – click this button and type/paste your organization’s URL to make your homepage the background behind the chat panel 
  • Copy link to share – click this button to copy the Preview URL to your clipboard, then forward it to your team 
  • Customize web chat – this button takes you to the web chat editor where you can personalize your chat panel’s elements 

Your colleagues are your best resource for refining your assistant’s content: they know the business, your customers’ pain points, and what questions most frequently need answering.  

Best practices we’ve identified for testing your assistant: 

  • Send the preview link to 10-15 colleagues 
  • Don’t invite anybody who was part of the building process  
  • Ask your colleagues to spend 10 minutes interacting with the assistant 
  • Ask them to log and categorize any issues they had 
  • Instruct your testers to ask for a human agent if they get frustrated  
  • Make sure you collect at least 20-30 conversations 

When reviewing conversation logs and your colleagues’ notes, the two main areas to focus on are:  

  • Understanding: Is your assistant properly understanding user requests? Are there frequent topics that you haven’t trained your assistant to handle yet?  
  • Resolution: Are users successfully completing actions? Is the assistant escalating actions to human agents more frequently for certain actions? 

You can review the details of each conversation in the Conversations tab of your assistant’s Analyze page (we’ll give you the tour of that page later). Reviewing conversations is an ideal way to improve your assistant’s actions. Once you deploy your assistant to your customers reviewing every conversation will be unrealistic, so take advantage of this early testing phase to rigorously refine your assistant’s actions. 

Customize the web chat

The web chat channel integration is automatically included with every instance of watsonx Assistant. Click Customize web chat on the Preview page or select Web chat on the home page to open the web chat editor.

Style, the first tab on the page, lets you name your assistant, adjust the text bubble and header colors, and add your company’s avatar (you’ll need to upload your avatar image to a URL to do this).  

Home Screen is where you can edit your assistant’s initial Greeting to the user. As the page states, “Good greetings are welcoming, actionable, and expressive of your assistant’s personality.”  

Once you’ve got a greeting that you’re happy with, add Conversation starters in the field below. These conversation starters should be simple phrases that match your assistant’s actions. They appear in the chat panel above the user input field.

 

When the user selects an option from among the starters, your assistant automatically funnels the user into the appropriate action — no heavy lifting required!

The web chat editor is also where you configure Suggestions. Suggestions are automatically configured on when you create your assistant. Suggestions appear in the chat panel when your assistant receives a request it doesn’t recognize. They’re designed to funnel the user into an action, just like the Conversation starters. 

The option label is preloaded with Connect to an agent. This option appears when the user clicks on the ? icon in the corner, or after three failed attempts (refer to Part II for more info on this feature).  

Get acquainted with your assistant’s analytics

Your assistant collects data from conversations with users in both the draft and the live environments. Once customers start interacting with your assistant, the Analyze page starts capturing data. This data will guide you as you continuously refine your assistant based on its success in resolving users’ requests. 

Select Get acquainted with your assistant’s analytics in the Task Tracker or click the chart icon to go to Analyze. 

You have the option to view analytics from a fixed date range (just open the dropdown menu and select a range) or to view a custom range based on specific dates.  

The centerpiece of the Analyze page is the Completion and Recognition tables. The data in these tables show you how often your assistant can successfully guide users through the end of an action, and how often it recognizes user requests. Beneath the tables, you’ll see lists of most frequent actions, least frequent actions, and least completed actions. 

The Conversations section lets you view actual interactions between end users and the assistant. You won’t need to guess what’s working and what isn’t, because you’ll have data tracking built into your assistant’s payload, plus real-world examples of interactions between your assistant and its users. 

Once you’ve taken the tour of the analytics and you’re ready to make use of it, take a moment to congratulate yourself: you’ve reached “the end of the beginning” and have an assistant that’s ready to meet your customers! 

Deploy your assistant

There are two short steps left before your customers can chat with your assistant on your website:

1. Publish your latest content and assign it to the live environment

Return to the Publish page by clicking the rocket icon on the side menu, or by clicking Publish your latest content to the live environment in the Task Tracker. You’ll see a list of all the edits you’ve made since publishing V1 in the table. Click the blue Publish button to publish V2 of your content. 

Note: We cover publishing, the two-environment structure, and lifecycle management in more depth in Getting Started Part IV. 

To assign V2 to the live environment, go to your Environments page, click Assign a version, then select V2 in the modal. 

Again, every time you publish a version of your assistant, you create a snapshot of your current content. The version assigned to the live environment will be the one your customers interact with, and any new content you add to your assistant will be saved in the draft environment until you’re ready to publish it as a version.

2. Deploy your assistant on a live channel across a broader set of customers.

When you’re ready to launch your assistant on your website, it’s literally as easy as copy and paste! Return to the web chat editor and click the Embed tab. 

You’ll see the JavaScript code snippet you’ll need to integrate your assistant into your website. Click the copy icon next to the script, open the HTML source for any page on your website, and paste the snippet in.  

Tip: Paste the snippet as close to the closing </body> tag as possible to ensure that your page renders as fast as possible. Refer to the documentation for watsonx Assistant to get the deepest dive possible on deploying your assistant to your website. 

Not only is your assistant now live and ready to start answering customers’ questions, you also have the option of deploying your assistant on as many of your site’s pages as you want. Just make sure you only add the code snippet once per page. 

Success! Now the fun begins

You now have a virtual agent live on your website, built, tested, and deployed in less than an hour!  

As great as this first version of your assistant is, you’ve only scratched the surface of what it can do. In Part IV we’ll cover the “Build – Deploy – Improve” lifecycle loop you’ll use to ensure that your assistant and its content are continuously becoming the best they can be. 

In the meantime, familiarize yourself with the Learning Center on your Assistant’s homepage to see all of our tutorial content, including orientation videos and our comprehensive product documentation. 

Be sure to follow IBM watsonx Assistant on Medium where we cover integrating your assistant, UX best practices, and generally how to make your customers’ experience a delightful one! That’s right, they’re going to get everything they need from your assistant, and they’re going to have fun doing it.

The final part of our Getting Started series covers your assistant’s lifecycle management. Read on! 

Was this article helpful?
YesNo

More from Artificial intelligence

What you need to know about the CCPA rules on AI and automated decision-making technology

9 min read - In November 2023, the California Privacy Protection Agency (CPPA) released a set of draft regulations on the use of artificial intelligence (AI) and automated decision-making technology (ADMT).  The proposed rules are still in development, but organizations may want to pay close attention to their evolution. Because the state is home to many of the world's biggest technology companies, any AI regulations that California adopts could have an impact far beyond its borders.  Furthermore, a California appeals court recently ruled that…

AI transforms the IT support experience

5 min read - We know that understanding clients’ technical issues is paramount for delivering effective support service. Enterprises demand prompt and accurate solutions to their technical issues, requiring support teams to possess deep technical knowledge and communicate action plans clearly. Product-embedded or online support tools, such as virtual assistants, can drive more informed and efficient support interactions with client self-service. About 85% of execs say generative AI will be interacting directly with customers in the next two years. Those who implement self-service search…

Bigger isn’t always better: How hybrid AI pattern enables smaller language models

5 min read - As large language models (LLMs) have entered the common vernacular, people have discovered how to use apps that access them. Modern AI tools can generate, create, summarize, translate, classify and even converse. Tools in the generative AI domain allow us to generate responses to prompts after learning from existing artifacts. One area that has not seen much innovation is at the far edge and on constrained devices. We see some versions of AI apps running locally on mobile devices with…

IBM Newsletters

Get our newsletters and topic updates that deliver the latest thought leadership and insights on emerging trends.
Subscribe now More newsletters