Bot Games Season 4 Generative AI Edition Solution Tutorial Challenge #1

Bot Games Season 4 Generative AI Edition Solution Tutorial Challenge #1
Userlevel 5
Badge +8

Challenge Page Tutorial - Healthcare AI

 

 

Bot Games Challenge Pages provide an excellent opportunity to learn about automation development. Have you completed the Healthcare AI Challenge? If you have, that's fantastic! If not, don't worry - we're here to show you one approach to solve this challenge and discuss other methods to help you automate this use case. Watch the video above for a complete walk-through of the challenge solution. Additionally, we're outlining some pointers below to help you maximize your speed if you want to see how quickly you can finish the challenge.

Here is the link to the Healthcare AI Challenge Page with full details.
 

Step Package

 

If looking at an entire end-to-end process feels overwhelming, consider first thinking through the major steps required to complete the task, then work on each step, one at a time. In the tutorial video above, we break down the entire bot into multiple steps. Using steps can help developers create logical groupings/separations in their code, as well as provide a basic "skeleton" of logic so they know exactly what actions need to be added to each section. Separate steps also improve the readability of your code in the case that you're working alongside other developers, or are showing the bot to business users who may not need to see every detail of the bot's operations.
 

Browser Package

 

To launch web applications, you can use the Browser package. The Browser package is also great for downloading files that are made available as links on web applications. You can use the ‘Download File’ action to download files without even having to even open the browser UI!
 

Recorder Package

 

With this package, you can interact with both browsers and windows applications. This package can work with multiple browsers in the same automation and can differentiate between multiple web applications based on the window titles.
 

OpenAI Generative AI Package

 

This cutting-edge package brings the power of OpenAI's advanced generative AI technology directly to the Automation Success Platform. It functions with GPT-4, ChatGPT-3.5 Turbo, and GPT-3 models with both OpenAI and Microsoft Azure service providers. Check out this link to learn more about this package and its capabilities.
 

JSON package

 

This package can help extract the required information from JSON text or files and use the extracted values in your automations. Check out this link to learn more about this package.
 

List Package

 

The List package contains actions that enable you to perform various operations on a variable of the list data type. In this use case, we have used the action “Size” to retrieve the number of items in a list and assign the output to a number variable.
 

Alternative Solutions

 

The tutorial above shows one of the most straightforward ways to complete this challenge, especially for those without extensive programming/scripting backgrounds. Should you want to explore the use of some alternate approaches to investigate their impact on processing times, here are a few modifications you could try:
 

  • Explore different AI models in ChatGPT to see which one performs better. Try different temperature values to ensure the response is always predictable. Keep it within a range of 0.3 to 0.5.
  • Be specific when asking about the response format you need, as ChatGPT can respond in different formats such as JSON or dictionary.
  • Use a Bot Store package
    • There are several helpful packages in Bot Store that enable developers to automate processes more quickly. One in particular that may be of value to this exercise is the Web Automation Package. This package uses Selenium to automate web interfaces and enables developers to quickly read/modify objects on a webpage. If you're not familiar with this package, check out this helpful Bot Store spotlight video demonstrating how it can be used to solve the Customer Onboarding challenge.

 

Conclusion:

 

An effective way to incorporate Generative AI into intelligent automation is to extract and summarize data into a standardized format. Generative AI can be utilized to extract specific information from long-form unstructured text, which your automations can then leverage to automate complex end-to-end workflows. This technique can be highly beneficial in various domains, such as extracting information from legal documents, research papers, and other documents with unstructured data. By implementing Generative AI with automation, you can improve the accuracy and efficiency of your workflows, while reducing manual labor and minimizing errors. With the vast amount of unstructured data that exists today, the potential for this approach is virtually limitless.

Also, one of the enjoyable aspects of building automations is discovering the various methods of solving the same problem. You can try using different packages or actions to complete the challenge and observe how those changes affect your automation’s processing time, accuracy, and stability.
 

Unlock all Bot Games Season 4 challenges and resources - visit our Bot Games Season 4 HQ today! 


0 replies

Be the first to reply!

Reply