AWS Step-by-Step
Using Amazon Quick to Build Your Own Enterprise Chatbot, Part 2
In the first part of this article series, I began walking you through the process of creating your own enterprise chatbot. If you have followed my instructions so far, then you likely have a functional chatbot, but the problem is that the chatbot is limited to using general knowledge. For the chatbot to be truly useful, it needs to be able to access your data.
Before I show you how to connect the chatbot to your data, I want to take a moment and talk about a couple more ways that you can customize your enterprise chatbot. The most obvious option is to modify the agent persona, but there are other options. For starters, you can expand the Customization section, found at the bottom of the preview screen. This causes Amazon Quick to expose options that you can use to customize the chatbot's welcome prompt as well as the prompt suggestions. Imagine, for example, that you were creating a chatbot designed to allow employees to ask questions about the organization's policies. You might change one of the suggested prompts from "tell me something interesting" to "how much vacation time do I get?" You can see the customization options shown in Figure 1.
[Click on image for larger view.] Figure 1: You Can Customize the Chatbot's Welcome Message and the Suggested Prompts.
Another way that you can customize the chatbot's behavior is by attaching reference documents. If you look at the previous screen capture, you can see that the interface includes a section for attaching reference documents. A reference document is a text file containing response templates, workflows, methodologies, or examples. You can use these documents to guide the chatbot's behavior. Reference documents can be in any commonly used format (such as .TXT, .PDF, .HTML, .DOC, .DOCX, .md, or .CSV), and each reference document can include up to 100,000 characters.
Now that I have talked about the various customization options available, let's talk about how to link your chatbot to your enterprise data.
One option is to link your chatbot to data sources that are outside of your AWS account. As an example, you might provide the chatbot with access to data residing within a SaaS application.
The exact steps involved in providing the chatbot with access to application data vary depending on the application. In any case, you would locate the console's Actions section and then click the Create button. This causes Amazon Quick to display a screen listing various integrations, as shown in Figure 2. Simply choose an integration and then follow the prompts to create an action. Once you have created the action, you can use the Link Action button, shown within the main console screen, to connect the chatbot to the action that you have created.
[Click on image for larger view.] Figure 2: Amazon Provides Numerous Integrations That You Can Use to Link a Chatbot to an External Data Source.
Your other option for connecting a chatbot to a data source is to create a knowledge source. To do so, locate the console's Knowledge Sources section and click the Create button. This brings you to the knowledge source screen, shown in Figure 3.
[Click on image for larger view.] Figure 3: You Can Create a Custom Knowledge Source for Your Chatbot to Use.
As its name suggests, a knowledge source is a collection of knowledge. The thing that makes knowledge sources so powerful is that they can contain several different types of knowledge. The easiest option is to provide a name for your knowledge source, select the File uploads tab, upload some files, and then click Share. This is a great option for situations where you need to build a chatbot that can answer questions about one or more documents (such as an employee handbook or a company calendar). However, your options do not end there.
If you look at the various tabs shown in the screen capture, you will notice that a knowledge source can include one or more actions. This means that you could conceivably build a knowledge source that contains files you have uploaded and data residing within external SaaS applications. A knowledge source can also include dashboards, knowledge bases, and topics. You can create these elements from the main Amazon Quick screen.
Once you have created a knowledge source, you can link it to your chatbot by clicking the Link Spaces button. When you are done and have tested your chatbot to your satisfaction, you can publish it by clicking the Launch Chat Agent button.
About the Author
Brien Posey is a 22-time Microsoft MVP with decades of IT experience. As a freelance writer, Posey has written thousands of articles and contributed to several dozen books on a wide variety of IT topics. Prior to going freelance, Posey was a CIO for a national chain of hospitals and health care facilities. He has also served as a network administrator for some of the country's largest insurance companies and for the Department of Defense at Fort Knox. In addition to his continued work in IT, Posey has spent the last several years actively training as a commercial scientist-astronaut candidate in preparation to fly on a mission to study polar mesospheric clouds from space. You can follow his spaceflight training on his Web site.