Yours2Learn

Inspiration

E-learning can be less engaging than in-person learning, yet it has the potential to provide even better, personalized learning opportunities. As high school students required to take two mandatory e-learning courses, my team and I experienced firsthand the challenges of staying focused and motivated. We wanted to create a tool that makes online learning more adaptive, interactive, and effective for students like us.

What it does

Yours2Learn is an AI-powered study tool that uses a webcam to track student focus during e-learning. The platform runs multiple AI agents to analyze engagement and learning patterns, then determines which agent is best for each task—such as creating lessons, generating quizzes, or giving study recommendations. By matching the right AI approach to each learning task, Yours2Learn delivers personalized, adaptive content that improves efficiency, keeps students engaged, and supports better learning outcomes.

How we built it

Teachers create learning tasks by defining the topic, purpose, and length. These tasks are routed through Solace Agent Mesh, which orchestrates multiple AI models (provided by us) to generate different versions of the same lesson for different student groups.

As students engage with the content, Presage analyzes webcam-based signals (such as eye movement and attention stability) to infer engagement and fatigue in real time. These signals are converted into structured feedback scores rather than storing raw data.

We store these feedback scores in MongoDB Atlas and use them to evaluate which AI models and content strategies perform best for specific learning goals. Over time, the system adapts by prioritizing AI configurations that consistently produce higher engagement.

The frontend is built with React Native, while Sentry is used to track critical learning failures (e.g. sudden engagement drops) rather than traditional application errors.

Challenges we ran into

We faced several obstacles along the way:

1. Presage not running properly on Mac → we learned to use Docker and even downloaded it from source to make it work.

2. Solace was difficult to configure and use, adding complexity to our workflow.

3. Limited team availability → as a team of two high schoolers, we had to balance building Yours2Learn and exam season

Through these challenges, we learned practical technical skills like Docker usage, troubleshooting complex software, and working under tight deadlines. More importantly, we learned how to approach real-world problems as students, designing solutions that actually meet the needs of our peers.

Accomplishments that we're proud of

1. Got Presage running successfully, overcoming technical issues on Mac.

2. Collected high-quality engagement data from our tests.

3. Built a system that effectively compares multiple AI agents to determine the best approach for each learning task.

4. Created a tool that can genuinely help students, like ourselves, optimize our online learning experience

What we learned

1. Learned how to set up and use Presage for focus tracking.

2. Gained experience with Sentry for monitoring and analyzing AI agent performance.

3. Learned to compare multiple AI agents and determine which is best for specific tasks.

4. Developed skills in debugging and troubleshooting complex software under technical constraints.

5. Learned to collect and process high-quality engagement data for adaptive learning.

What's next for Yours2Learn

1. Expand AI-comparison metrics to evaluate AI generation for specific subjects, lab exercises, videos, and other interactive content.

2. Create a video generation option alongside existing quizzes and slides to make lessons richer and more engaging.

Built With

• css
• github
• javascript
• mongodb
• presage
• python
• react
• sentry
• surveymonkey