AI-Powered Scenario eLearning
Proof of Concept, 2025 - 2026
The Challenge
Traditional eLearning presents static scenarios with predetermined feedback. Learners click through branching paths, but responses are generic and cannot adapt to what someone actually writes. My dissertation research showed that scenario-based approaches surface real reasoning gaps, but existing tools could not respond to that reasoning in the moment.
I designed and built an interactive scenario-based training module that uses generative AI to provide real-time, personalized feedback on learner responses. Built with React and the Claude API, it presents realistic workplace scenarios and evaluates learner reasoning against defined competency criteria. This proof of concept translates my dissertation findings into a working tool.
What I Built
- Designed scenario structure using the CII framework (Concern, Impact, Invitation) for supervisory communication
- Built a React front end with component-based architecture
- Integrated the Claude API for real-time response evaluation
- Crafted system prompts designed as assessment rubrics tied to learning objectives
- Connected design decisions directly to dissertation research on scenario-based learning transfer
- Built responsive design for desktop and mobile
Try the Demo
This demo uses pre-written coaching responses to simulate the AI feedback experience. The production version connects to Claude for personalized feedback.
Launch Interactive DemoResults
- Demonstrated that generative AI can create adaptive learning experiences
- Built a reusable architecture applicable to any professional development topic
- Translated dissertation research findings into a working prototype
- Created a model for AI feedback that supplements, not replaces, human facilitation
Connect
I welcome conversations about educational leadership, learning strategy, professional development systems, research, speaking, and consulting.