FLASHCRDS

• 

  screens

Inspiration

What sparked this idea was my current role as a programming tutor for K-6 kids. I wanted to address a common challenge faced by teachers every day – how to best educate children. Ensuring that education for kids is as effective as possible is crucial, especially considering the significant negative impact the pandemic had on children during some of their most developmental years. I aimed to create something that could potentially alleviate this issue. What I've learned is that children respond well to gamified learning, and since they enjoy quiz games like Kahoot, I thought a similar approach with flashcards could be beneficial and very effective.

What it does

FLASHCRDS, powered by Gemini, works by converting a textbook PDF into raw text. Gemini processes this text input and generates intelligent flashcards segmented by chapter.

How we built it

FLASHCRDS is a full-stack web application developed using React, TypeScript, Next.js, Tailwind CSS, and ShadcnUI. User authentication is handled through Clerk.

Challenges we ran into

Initially, I implemented a file upload system for teachers to upload PDFs directly, but integrating Gemini AI with file uploads posed difficulties. Gemini couldn't directly handle PDFs, which disrupted the development process. Consequently, I had to manually extract text from PDFs and feed it into Gemini to generate flashcards, instead of Gemini directly reading the PDFs.

Accomplishments that we're proud of

I'm proud of the app's idea, as I believe it addresses a real-world need and could have a significant impact alongside its ability to scale with further resources. Leveraging a new API and gaining insights into AI during the project were also notable achievements. Lastly, I'm satisfied with the visual design of the app; I think it looks fantastic.

What we learned

Throughout this project, I gained valuable insights into AI and its applications in education, as well as honed my skills in web development, particularly with backend calls to the Gemini API. I also learned a lot about the use cases for AI and especially Gemini.

What's next for FLASHCRDS

Flashcards is far from a finished product. There's still much to do, including implementing a database connection, authentication, and authorization for teachers and user profiles for students. Additionally, refining the generative AI aspect of the app and reintroducing the file upload feature are priorities. I envision further improvements to Gemini, such as the implementation of direct PDF processing, which would eliminate the need for intermediary text conversion.

Built With

• gemini
• next.js
• node.js
• react
• shadcn
• tailwind
• typescript