This hands-on lab provides an Azure administrator with an understanding of how to implement a simple Web App Bot within your organization’s IT services. Web App Bots can be used to enrich customer or employee experience as they interact with web-based services. You will learn how to implement a Bot within Azure that can be used by customers or employees to answer common questions and minimize the overhead cost of repetitive or common tasks.
Understand the scenario
You are an AI consultant for a data and analytics vendor. You need to create and configure an Azure Web App Bot that will answer common questions from your user base. First, you will create an Azure App Service plan. Next, you will create a bot by using the Azure Bot Service. Finally, you will test the bot, and then you will review performance metrics for the bot.
Create an Azure App Service plan:
You will first implement a new App Services Plan. The App Services Plan is essentially a container with a specific capacity, where you can host multiple App Services. For this lesson, you will create an App Service Plan and learn the basics of configuring your allocated compute resources for the plan (e.g., region, base operating system, number and size of VM instances, and pricing tiers). Setting up an App Service Plan is the cloud-based equivalent of standing up an on-premise server farm for hosting your web applications.
Create an Azure Web App bot:
Microsoft provides a framework for creating enterprise-grade bots that leverage cognitive services, such as AI and machine learning. The Azure Web App Bot enables you to build intelligent bots while maintaining control of your data. These bots use Language Understanding (LUIS) service to recognize natural language and predict an understanding of what is asked, then to offer an automated response to the user’s needs. LUIS interprets your language inputs, or utterances in the form of chat statements or questions, evaluates them, and returns results based on what LUIS predicts is your intent. In this section, you will learn how to configure a basic Web App Bot, which provides you with a message endpoint that can be added to your Web Apps.
Test the Web App Bot:
For this task, you will test the functionality of the Web App Bot within the Azure Portal. This feature is great for troubleshooting your bot, to ensure that it is operating properly and eliminates variables that are introduced, depending on how the bot is deployed. You will see how the bot responds to your natural language input, predicts what your intent is, and helps you accomplish a simple task. This capability provides a two-fold value to an organization by first enriching the user experience and secondly by minimizing the overhead created by handling the most common tasks through AI and automation.
Verify the performance of the bot:
As the administrator of your Azure environment and these valuable capabilities, you need to be able to troubleshoot and monitor the performance of these services. In this section, you will learn how to evaluate the performance of your bot and ensure that it is functioning properly. You will learn about the activity log and evaluate the performance of the bot to make sure there are no critical or error events. It is also worth considering how you could automate notification of issues, leveraging the Azure platform. For example, implementation of log analytics and establishing messaging services to send notifications if your bot service fails.
Lab Summary Conclusion:
In this hands-on virtual lab, you will gain valuable insights into Microsoft Azure Web App Bot service. This service can bring tremendous value to your existing web app services by giving customers an easy to use chat service to automatically respond to their most common questions or issues. It can also drastically reduce the overhead that would otherwise be required for a human to respond to the same problems. Using Azure’s cognitive services, which are AI and machine learning-based, are essential skills for someone pursuing a career as an Azure administrator, especially in the field of automation and data science.
Other Challenges in this series
- GUIDED CHALLENGE: Configure and Query the Cognitive Services Face API
- ADVANCED CHALLENGE: Can You Provision and Analyze a Cognitive Services Computer Vision API?