RecallRadar — exam revision on Bee

Inspiration

What it does

How we built it

Challenges we ran into

Accomplishments that we're proud of

What we learned

What's next for RecallRadar — exam revision on Bee

Inspiration

We noticed that students often struggle to filter vast amounts of study material into actionable revision sessions. We wanted to create a tool that uses advanced machine learning to transform static notes into active recall exercises automatically. Our goal was to bridge the gap between passive reading and active testing using the Bee ecosystem.

What it does

RecallRadar allows users to upload course materials and receive a comprehensive exam revision plan instantly. The platform uses AI to identify key concepts and generates mock tests that simulate real exam environments. Users can track their progress over time to ensure they are fully prepared for their final assessments.

How we built it

We built the backend using FastAPI and deployed our services on Amazon ECS Fargate for scalability and reliability. We integrated Amazon SageMaker to host the models that process study text and generate intelligent questions. The entire architecture is optimized for the Bee track to ensure a seamless experience for all developers.

Challenges we ran into

One major challenge was optimizing the SageMaker inference time to ensure low latency during live study sessions. We also struggled initially with configuring the networking between our ECS Fargate containers and the Bee surface APIs. We resolved these issues by refining our VPC configurations and implementing asynchronous task processing.\n## Accomplishments that we're proud of We are incredibly proud of how seamlessly the FastAPI backend integrates with the SageMaker models to provide instant feedback. The accuracy of the generated exam questions exceeded our initial expectations during the testing phase. Successfully deploying a robust architecture on the Bee track was also a major milestone for our team.

What we learned

We learned a lot about managing containerized applications using Amazon ECS Fargate in a production environment. We also gained deep insights into how to structure machine learning workflows specifically for educational use-cases. Understanding the unique constraints and opportunities of the Bee surface helped us design a much better user interface.

What's next for RecallRadar

We plan to implement a collaborative study mode where multiple users can revise together on the Bee platform. We also want to expand our supported file formats to include video lectures and handwritten notes. Future updates will focus on predictive analytics to forecast exam scores based on revision performance data.

Built With

• amazon-bee
• amazon-ecs-fargate
• exam-revision
• fastapi
• recallradar
• sagemaker