Memory Hammer

Memory Hammer helps you create flash cards for anything using Google AI and remember it using spaced-repetition

Comment

Inspiration

After seeing many friends and relatives having issues with their memory at young age, especially after COVID induced brain fog; I wanted to create a tool which helps them and anyone to augment their long-term memory through spaced-repetition.

What it does

Spaced repetition is an evidence-based learning technique that is usually performed with flashcards. Newly introduced and more difficult flashcards are shown more frequently, while older and less difficult flashcards are shown less frequently in order to exploit the psychological spacing effect. - Wikipedia.

But creating flashcards in the current spaced-repetition apps are tedious since it's manual, Memory Hammer creates flashcards automatically for anything we want to remember using Google AI and enables spaced-repetition learning using FSRS algorithm.

Text to Flashcards

Enter, copy/paste or share text from any other apps to Memory Hammer.

Text box with text content and a submit button below it

Flashcards are created automatically from the text using Google AI (Gemini)

Preview Flashcards

Preview, Save and Export flashcards generated from the given text

Front

Front side of the flashcard

Back

Back side of the flashcard

Image to Flashcards

Capture photo/pick from gallery/share image from any other apps to Memory Hammer

Screen showing buttons to capture photo or pick image from gallery

Capture photo

Photo capturing screen

Flashcards are created automatically from the images using Google AI (Gemini)

Preview, Save and Export flashcards generated from the given photo/image

Front

Front screen of the flashcard

Back

Back screen of the flashcard

Search the flashcards

Search flashcards

Search screen for the flashcards

Delete flashcard

Delete dialog for the flashcard

Review Flashcards

Review flashcards by clicking the top right review star button, Choose a rating for the card - Again, Hard, Good, Easy depending upon how well you remember the card; The next due is scheduled automatically.

Front screen for the review flashcard

Back screen for the review flashcard

Dark mode

screen showing dark mode theme

Testing

Link to private GitHub repository has been submitted in the additional info section. Please login to GitHub with testing@devpost.com to access the repository and follow the Readme for testing instructions.

How I built it

I built it using Flutter Gemini API as I want Memory Hammer to be a cross-platform application(currently supports android), I use SQLite3 for database.

I studied the need-gap in memory health apps, conceptualized the idea, Implemented UI/UX, database and integrated the FSRS algorithm. Tested and improved the UI/performance of the application over the course of the competition.

Note: I secure the Gemini API key by asking the user to enter the API key themselves. When deploying it to Playstore/Appstore, I will move the API calls to the back-end.

API key dialog in memory hammer

Challenges we ran into

It wasn't clear on how I can use gemini-1.5-pro-latest model in the flutter API, So I had to dig into the API code and found that I had to pass apiVersion as v1beta through RequestOptions in the requests. Discussions by others in the forum on the same helped me find the solution.

There were other minor difficulties along the way of developing Memory Hammer, But Google AI/ Gemini's robust API documentation and guides helped me clear every one of them without much hassle.

Accomplishments that we're proud of

I'm proud of being able to create Memory Hammer to solve memory health problems using Google AI within the stipulated period of the hackathon.

What we learned

I learnt about Google AI and Gemini API, improved my skills with flutter. Gained knowledge in spaced-repetition technique.

What's next for Memory Hammer

I want to release Memory Hammer as a production application with Gemini API in the back-end.

Built With

  • ai
  • dart
  • flutter
  • gemini
  • google-ai
  • health
  • memory
Share this project:

Updates