The Problem
Generative AI is increasingly being used to support lesson planning. It can help teachers save time and quickly generate diverse lesson ideas. However, the quality of AI-generated lesson plans can depend heavily on how teachers structure their prompts, and the initial output may not always reflect the needs of a specific classroom or instructional context.
LessonAI addresses this challenge by providing AI-supported lesson generation while keeping teachers actively involved in the design process. Teachers can edit generated lesson plans based on their instructional needs and use an evidence-based evaluation rubric to assess lesson quality. By combining AI support with teacher judgment, LessonAI is designed to help educators save time while maintaining the quality of their instructional decisions.
Design Challenge
How might we use AI to make lesson planning more efficient while keeping teachers in control of instructional quality and decision-making?
Faculty Collaboration & Design Process
I collaborated with faculty and educators throughout the development process, using continuous communication to translate instructional needs into product decisions. Rather than developing the platform independently and asking for feedback only at the end, I involved stakeholders throughout ideation, prototyping, development, and testing.
Ideation & Design Intent
Discussed the initial concept with stakeholders to understand their instructional needs, expectations, and the intended role of AI in lesson planning.
Prototype & Alignment
Built early prototypes and shared them with stakeholders to confirm that the proposed design reflected their instructional intentions before moving into full development.
Product Development
Developed and refined the platform based on prototype feedback, translating educator needs into product features and interaction decisions.
User Testing & Iteration
Tested the developing product with educators, identified usability and instructional-design issues, and iterated on the experience based on their feedback.
My Role: Learning Design • Product Design • AI Prompting • Frontend Development • User Testing & Iteration
Design Principle
Research provides the foundation; educator expertise provides the context.
The platform combines research-informed design with continuous educator feedback so that AI supports real instructional needs rather than determining instructional decisions.
Research-Informed Lesson Evaluation
The evaluator uses a rubric developed by faculty and grounded in instructional design research. It reviews major elements of lesson quality, including the clarity of the lesson focus, alignment among objectives, standards, activities, and assessment, support for diverse learners, pacing, resources, and opportunities for reflection.
The detailed scoring language is intentionally not displayed in this portfolio. Instead, this page focuses on how the rubric supports the product experience and teacher decision-making.
Teacher Agency Through Evaluation Override
AI evaluation is used as a starting point rather than a final judgment. Teachers can review the AI rating, compare it with their own professional judgment, override the evaluation, and document additional context. This interaction is designed to support AI literacy by making evaluation transparent, editable, and open to disagreement.
Review the AI Evaluation
The teacher sees an AI-generated rating and a short explanation of the reasoning behind it.
Apply Professional Judgment
The teacher can choose a different rating when the AI does not reflect the lesson context or instructional intention.
Add Context and Improve the Lesson
Teacher notes capture contextual information that the AI may not understand and can guide later revisions.
Customizable Lesson Generation
Before generating a lesson, teachers can select lesson-plan preferences that reflect their classroom needs. These options allow the generated first draft to better match the teacher's intended instructional approach instead of producing a generic lesson.
Select Lesson Preferences
Teachers choose options such as technology usage, instructional approach, teaching strategy, and other advanced lesson settings.
Generate a Contextualized First Draft
The system uses the selected preferences to generate a lesson plan that better reflects the teacher's goals.
Evaluate and Revise
The teacher can evaluate the generated lesson, revise the content, and make the final instructional decision.
Explore the Product
The live platform allows users to explore the lesson generator, advanced preference settings, lesson evaluation workflow, and teacher override features.
The portfolio describes the purpose and structure of the faculty-developed rubric without reproducing its full scoring criteria.