MemorAI – AI supercharged flashcards

Inspiration

As college students, we were seeking more efficient ways of studying so we looked to apps such as Quizlet and Anki ad inspiration. We decided to create our own flashcard web application with a personalized AI algorithm to optimize memorization with the user-friendliness of Quizlet and without the excessiveness of Anki.

What it does

MemorAI uses an algorithm for personalized learning which measures how well users recall certain concepts and tailor optimal repetitions for each deck, collecting data for progress reports improvement of accuracy.

How we built it

Using Figma, we designed a visually pleasing, user-friendly interface and implemented it with React. We build the backend using Google Cloud Functions in Firebase and stored data in Google Cloud Base. We trained our models by using Kubernetes docker containers.

Challenges we ran into

The biggest challenge was integrating various languages and components into a cohesive project. Another challenge was learning the theoretical concepts of machine learning and finding good sources to base our model on.

Accomplishments that we're proud of

algorITHIM

What we learned

We learned to work as a team, specifically how to delegate tasks for not only maximum efficiency but so that each person can learn new skills while contributing. One improvement from our first hackathon was our spec development; by making sure each member understood how their individual component fits into the project, ensuring clearer insight and objectives of the final product. We also learned how to not only read research papers but also find good research papers to draw inspiration from by exploring Archive.

What's next for MemorAI?

We will continue to work on refining the algorithm and improving the infrastructure and launch a beta test for our friends and peers.

Built With

• figma
• firebase
• google-cloud
• node.js
• react
• tensorflow