AWS Step-by-Step
Using Amazon Quick to Build Your Own Enterprise Chatbot, Part 1
Amazon Quick makes it surprisingly easy to build your own enterprise chatbot. Unlike general-purpose chatbots such as ChatGPT or Grok, your custom enterprise chatbot can be provided with your organization's own internal data. By doing so, you could conceivably build chatbots that help with the employee onboarding process or that help with project management. You could even build a chatbot that acts as a research assistant using your organization's own proprietary data. The potential use cases are endless, and all you need is an Amazon Quick account.
To get started, sign into your Amazon Quick account. From the home screen, click the Chat Agents tab and then click the Create Chat Agent button, shown in Figure 1.
[Click on image for larger view.] Figure 1: Click on the Create Chat Agent Button.
At this point, you will be taken to a screen that asks you to provide a description of what the chat agent should do. As you can see in Figure 2, the prompt box initially contains a template that is designed to help you through the agent creation process. The template initially says, "I want a chat agent that [does what] for [whom]. It should use knowledge about [topics] and be able to [specific capabilities]. The tone should be [style]."
[Click on image for larger view.] Figure 2: Amazon Provides a Template That You Can Use.
You can use this template as-is, simply replacing the text within the brackets. If you need some additional help, there are examples shown at the bottom of the screen.
The first example is, "Create an agent that helps new team members to learn about our processes and tools. Should answer common questions and guide people to relevant resources and documentation." Amazon's second example is, "Build an agent that helps track project status, find relevant documentation, and assist with common project tasks. Should understand project management basics and help with status updates." One final example provided by Amazon is, "Create an agent that guides people through setup and access requests. Should help with common system access needs, tool setup, and finding basic getting started documentation." It is worth noting that even though the examples provided by Amazon are short and to the point, your prompt can contain up to 50,000 characters.
When you have entered your initial prompt, click Generate. After a few seconds, AWS will take you to a screen, similar to the one shown in Figure 3, where you can fine-tune the chatbot that you are creating.
[Click on image for larger view.] Figure 3: This Screen Is Used to Fine-Tune the Chatbot That You Are Creating.
There is quite a bit to talk about with regard to this screen. I want to begin with the pane on the right. This portion of the interface displays a preview of the chatbot that you are creating. Here, you can ask questions and see how the chatbot responds. The important thing is to verify that the chatbot behaves in a way that aligns with the instructions you have provided.
There are a couple of sample prompts shown at the bottom of the screen, and you can use these samples if you are stuck for ideas. However, the main things I want to point out are the options that exist at the bottom of the chat interface.
The first of these options is a dropdown that says All Data. As you can see in Figure 4, you can click on this dropdown and then tell the chatbot to use general knowledge or specific data instead. The bottom of the chat window also features a globe icon you can select to allow the chatbot to retrieve information from the Internet. There is also a paperclip icon you can use to attach a file to your prompt.
[Click on image for larger view.] Figure 4: You Can Tell the Chatbot What Data to Use.
The opposite side of the chat window contains a Smart dropdown, which you can see in Figure 5. Clicking on this dropdown gives you the chance to choose whether you want the chatbot to respond quickly, whether it should take longer but use advanced reasoning, or if it should strike a balance between performance and speed.
[Click on image for larger view.] Figure 5: You Can Choose the Speed and Depth of the Request.
Now that I have shown you how to test the chatbot, I will continue the discussion in Part 2 by showing you how to connect your chatbot to various data sources.
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.