Chatbot as a Training Tool- A Critical Analysis!

After having setup the chatbot and using it as a preparation tool to appear for and pass the AWS Certified Cloud Practioner exam, I am now in a position to objectively review the efficacy of the chatbot.

Below are my thoughts:

What works in favor of a Chatbot?

1. Learning on the Go: A chatbot works on a simple user interface that can be accessed by the user across multiple devices- Laptop/Tablet/Mobile Phone etc. This provides user a flexibility to consume ‘concentrated’ nuggets of information as and when they have the time and the appetite.
2. Ease of Browsing Topics: A chatbot allows users to move away from the traditional way of accessing topics in a linear and structured format and instead allows an intelligent implementer to provide references from the main topic to sub-topics that are either related or may be confused with
3. Ease of adding new Topics: This is an obvious advantage to the implementer as it allows them to add new topics or improve the overall experience of the end-user with regards to the training material with minimum effort for implementers and minimal impact or down-time to end users
4. Linking to external sources of information: For information that is dynamic, chatbot allows users to access URLs instead of static content thereby increasing the overall accuracy of the training content

What can be made better?

1. No way for users to “highlight” important topics for last minute preparation: Most users would like to refer to some key points identified by them during their initial study so they can refer to the same before the exam. This shortcoming can be mitigated by customizing the chatbot to allow users to command the chatbot to create an FAQ section which can be referred to later. I will research this point and see if this is really a feasible option (hopefully a separate blog post on it sometime in the future :))
2. No way for users to test their preparation: After going through the content , users would like to test their knowledge by going through mock tests. I also felt the need for the same and ended up enrolling for external practice exams. The chatbot can be enhanced by setting up a practice exam with questions similar in content to the actual exam. Since this chatbot is based on IBM Watson Assistant, I will try to explore the ‘Actions’ Skills for this.
3. Static text sometimes does not explain concepts clearly: Some concepts especially which involve complex setups e.g. S3 bucket creation cannot be effectively explained via static text. This can be handled by embedding videos instead.
4. Challenge of setting up the chatbot and preparing for the exam simultaneously- While this is not a shortcoming of the bot but rather a learning that working on these two activities at the same time was rather time consuming and at times, the two objectives seem conflicting.

As my conclusion, I strongly advocate using chatbots as training tools with the caveat that their efficacy can be greatly improved by filling in the gaps that I encountered.