Integrating AI Chatbots in eLearning Courses

This eLearning course explores AI as an interactive tool—not just for content generation or automation. Learners practice guiding conversations with an AI chatbot, applying soft skills like active listening and questioning in realistic, client-facing scenarios.

Audience: Learners developing client communication and travel consultation skills.

Responsibilities: Instructional Design, eLearning Development, Chatbot Integration, Prompt Engineering, Scenario Design, Visual Design

Tools Used: Articulate Storyline, Adobe Illustrator, ClueLabs, ChatGPT

This project was created in 2025.

Experience the Project

The Problem

AI is often used behind the scenes in eLearning—to write content, generate images, or assist with course building. But what if we used AI differently? I wanted to explore how AI could be brought into the learning experience itself, not just support its creation.

Many learners struggle to apply new knowledge in realistic scenarios. I saw an opportunity to leverage AI as an interactive roleplay tool—one that lets learners practice skills in a more dynamic, lifelike way.

The Solution

I designed an eLearning course that uses an AI chatbot as a conversation partner. Instead of taking a quiz, learners practice in a simulated client interaction—applying skills like listening, questioning, and guiding decisions in real time.

The chatbot plays a client unsure of where to travel. It responds naturally, giving learners a safe but realistic space to build soft skills like active listening, curiosity, and conversational flow—essential for client-facing roles.

This project shows how AI can move beyond content creation to transform how learners apply knowledge in digital environments.

My Process

To build this course, I used ChatGPT as a Subject Matter Expert to identify the key skills a travel agent needs and to create the 'storyboard' of the course. I then wrote a detailed prompt to program the AI chatbot’s behavior and guide its responses.

Using a widget produced by ClueLabs, I built the chatbot interaction directly in Articulate Storyline, allowing for a realistic, open-ended conversation experience. I sourced imagery from Freepik and customized the visuals in Adobe Illustrator to match the course’s travel theme. From there, I brought everything together—designing, animating, and developing the full course in Storyline.

Prompt Engineering the Chatbot

To bring the AI conversation to life, I used ChatGPT to simulate a realistic client interaction. My goal was to create a chatbot that would challenge learners to apply active listening—dropping subtle hints about a secret destination while responding naturally to the learner’s questions.

Rather than relying on a static script, I wrote a detailed prompt that trained the chatbot to behave like a real, slightly indecisive travel client. It was designed to:

Stay vague at first and only reveal more when the learner asked thoughtful, active questions
Respond with short, natural-sounding phrases to keep the learner leading the conversation
Give less helpful—or even slightly annoyed—responses if the learner ignored previous clues

Behind the scenes, I went through multiple rounds of refinement—tweaking the prompt to get the right tone, pacing, and level of challenge. I also programmed the behavior so it would randomly choose from a list of destinations and maintain consistency throughout the conversation.

To see the full prompt, click here.

Results and Takeaways

Integrating a chatbot into this eLearning course proved to be an engaging and effective way to simulate real-life conversations. Unlike static quizzes or branching scenarios, this approach gave learners space to apply their knowledge dynamically—responding, adjusting, and practicing soft skills in real time.

That said, it also surfaced some of the current limitations of using AI in eLearning:

Consistency and control: While powerful, chatbots don’t always behave exactly as intended. Prompt design requires careful testing and ongoing tweaks to shape the learner experience.
Data visibility: One area for future development is tracking learner-chatbot conversations. If conversations could be exported (e.g., to Excel), it would open the door to better feedback, analysis, and continuous improvement.
Security and approval: Many LLMs used in these tools may not be approved for use in certain organizations. Ideally, this type of chatbot could be built with a company-approved language model.
Cost considerations: Every AI interaction comes with a usage fee, so scalability and budgeting must be taken into account—especially for large learning audiences.

Overall, this project reinforced the exciting potential of chatbots as roleplay tools within eLearning. With the right strategy and infrastructure, they can bring realistic practice into the learning experience—bridging the gap between knowledge and confident application.

Page updated

Google Sites