Boost productivity and collaboration with watsonx Assistant’s organizational features
Introducing important new features that allow you to group, filter, and find actions quickly to enhance your build process
*Watson Assistant has been rebranded to watsonx Assistant. Any reference to it in this article refers to its new name.
In today’s fast-paced digital landscape, efficient organization and streamlined workflows are crucial for success. As part of our commitment to empowering conversational AI development, Watson Assistant offers a range of organizational features that enhance productivity and collaboration. In this blog post, we will dive into three of them:
- Collections
- Filtering
- Step Titles
Collections — simplifying action organization
Efficiently organizing actions is essential for the effective development and maintenance of your virtual assistant. To aid in this, Watson Assistant has introduced collections, an organizational feature that simplifies this process. With collections, teams can group related actions together, making it easier to locate and manage a specific sets of tasks.
Imagine a customer support team responsible for different domains such as billing, technical support, and account management. By creating collections for each domain, team members can quickly find and work on the relevant actions within their area of expertise. This targeted approach saves time, enhances collaboration, and boosts overall productivity.
There are two ways to create a collection. After creation, it is denoted by a folder icon in the table.
- The first approach is by clicking on the New collection icon and entering a name, and clicking on Add.
- The second approach is by selecting existing actions and clicking on Add to new collection. This will move the actions into the newly created collections.
Since collections introduces a folder structure, we added breadcrumbs on top of the actions table. If the breadcrumb looks like the below screenshot, it indicates the user is looking at the Banking collection. Users can navigate back to the main table by clicking on Created by you.
To learn more about collections in Watson Assistant, explore the topic in our documentation.
Filtering — it’s all about saving time
Managing large-scale conversational AI projects with numerous interconnected actions can be daunting. Watson Assistant’s filtering feature offers a powerful solution to efficiently navigate and locate specific actions. By applying filters based on variables, extensions, dependencies, and more, authors can quickly narrow down the list to find the actions they need.
For chatbot authoring teams, filtering is a game-changer. Whether it is searching for actions that use a specific variable, a call out to a particular extension, or serving as dependencies for other actions, this feature saves valuable time and improves focus. Authors can now dedicate more energy to developing and updating critical components, and less time looking for them.
You can find the filters by clicking on the filter icon in the toolbar above the actions table. This will expand the filter section view.
Next, there are 4 filters that can be applied to the table. They are:
- By subaction: The table will be updated to show actions that trigger or call the selected subaction. This is useful in situations where the author is updating a particular action, but knows that one or more actions are calling this action (subaction). The subaction filter gives the author an instant view of all the places that may be impacted by an update to the subaction.
- By extension: When applied, the table shows all actions that use the selected extension. Users can choose Any to see all the actions that use at least one custom extension.
- By variable: The table shows all actions that use the selected variable.
- Show all subactions: The table shows all actions that are triggered, or called, by other actions.
There are also additional benefits to using action filters:
- Suppose a user wants to delete an action. This action can be a subaction, and therefore when deleted it can cause the calling action to fail. To prevent this, we added a safety measure to warn the user. In the example below, they can simply click on View all to see all the actions that can trigger Confirm an appointment.
- In the Variables table, users can see how many actions a variable is being used in. To get more information on the actions themselves, users in the example below can click on View all. When this is selected, users will navigate to the actions table with the variable filter applied.
To discover more about the filtering feature in Watson Assistant, please refer to our documentation.
Step titles — enhance your understanding of steps
In complex conversational flows, understanding the purpose of each step is paramount. Watson Assistant has introduced step titles, an organizational feature that empowers authors to assign meaningful titles to individual steps within a dialogue. This functionality enhances comprehension, simplifies maintenance and improves collaboration among team members.
With step titles, authors can quickly grasp the purpose of each step at a glance. From clarification prompts to data collection actions, meaningful titles provide a visual aid for organizing and comprehending complex dialogues. This feature ensures teams remain aligned and promotes efficiency throughout the development process. Historically Watson Assistant used the text of a given step as its pseudo-title. Now chatbot authors can quickly see the step’s title in the list of variables, and more easily understand the variables in the variable picker list.
In 2022, we added the ability for authors to provide a note (or description) of each action. In the context of the more recent organizational feature updates, action notes allow the author to provide details about the action’s behavior for other authors to get up to speed.
Conclusion
Organizational features are the backbone of efficient and productive conversational AI development. In this blog post, we explored three key features in Watson Assistant:
- Collections (to simplify action organization)
- Filtering (to save time by narrowing down search results)
- Step titles (to enhance assistant organization)
By harnessing the power of these features, you can unlock the full potential of Watson Assistant and streamline the conversational AI development processes. To learn more about these features and discover additional resources, visit the Watson Assistant documentation.
Embrace these organizational tools and elevate your virtual agent development to new heights!
Be sure to check out new routing features on Watson Assistant, as well as Watson Assistant learning.
