This project is still under NDA. Please come back another time :)
Building teacher dashboards using AI learning science tools
RightOn! Education ❋ Product Designer ❋ Jan - Mar (2026)
Overview
RightOn! is an online quiz game platform that helps classrooms turn mistakes into learning opportunities. Through its setup, it aims to destigmatize errors to create a more fun, social, and supportive learning environment.
I designed and validated the end to end AI game creation workflow that is now being built for classroom pilots. In the process, I created 3 new key components, 13 component states, and 20+ screens.
Toolkit: Figma, Prototyping, Gen AI
🧩 Challenge
Teachers find the question and game creation process
time-consuming.
Unlike most other quiz platforms that require just questions and answers, RightOn! has designed a specific experience based on pedagogical research that requires users to develop questions, solution steps, distractors, and explanations in order to create a game.
The team recognized an opportunity to incorporate AI into the game creation process, helping teachers generate questions, answers, and explanations while keeping teachers in control of quality and context.
How might we design an AI-powered experience to make the game creation process more efficient for teachers?
🚀 Discovery
Understanding how AI impacts workflow
RightOn has never worked with AI before, so I started by diving into research on how to design for generative AI experiences. Building upon the existing game creation flow, I ideated multiple user flows that integrated AI. I debated three question generation options with my team, ultimately landing on the flow that provided the most value to users while balancing human intervention.
My final flow is one that accounts for generative variability, trust & accuracy, and user co-creation with AI.
🎨 Design
How will teachers interact with AI?
Generative AI presents a new form of human-computer interaction, called intent-based outcome, that comes with its own set of challenges.
Prompt assistance helps users articulate their thoughts
The prompting component scaffolds users through key context inputs so the generated game aligns with their intent.
Presenting dense info in a useful way
I created 6 iterations of modal layouts to display generated questions, experimenting with the content to display, card states, and horizontal vs. vertical layouts.
The winning design (below) distilled all of this content into a compact layout without sacrificing clarity or usability.
Designing the AI feedback mechanism to improve output
To do this, I had to think like a dev. How does user behavior indicate to the algorithm what works and what doesn't?
📝 Testing & Feedback
Ensuring teachers understand how to use AI
Through pilot testing, two usability tests, and discussions with the PM and developers, we gained new insights for our next iteration. Teachers showed that they didn't have trouble understanding how to handle AI, but did lead us to minor UI and UX changes. Info was added or removed to reflect user needs and the preview panel became editable to allow teachers to make quick changes.
🤖 Future
Things don't always go as planned!
This sprint was only planned to be two weeks long, so we had to re-prioritize certain features for future sprints. The priority was shipping the AI quickly, so that teachers could begin using it.
We didn't have enough time to test the tags through a card sort, so we backlogged the feature. We still needed to understand what tags would be most useful to teachers and how to order them.
I originally proposed a chat log to let users reference previous prompts and outputs. User testing revealed a strong preference for chatbot interfaces, which aligned with their existing mental models. However, due to time constraints and the complexity of this feature, it also needed to be backlogged for future sprints.
Depicted: ChatGPT's chat log interface, which is how users are most familiar with interacting with AI.
🏆 Result
View my UI/UX designs on the website once the feature rolls out!
I designed 20+ screens.
Here's a snapshot of some of the key screens.
I added 3 new components to the design system.
Here's a snapshot of the components and their key states.
💭 Reflection
Real-world constraints mean processes are imperfect.
The final theme, "Where ideas take flight".
RightOn Education is a non-profit organization, meaning they have limited resources to conduct research and testing. Our team was only able to find two users within our target demographic (math teachers), but that was okay! We gained valuable insights that we wouldn't have been able to discover otherwise. It was crucial that we had fresh new eyes and a different perspective to evaluate our prototype.