TiCard

Home Page
Flashcard Page
Definition and Formula

Inspiration

As students, we want to make something that can solve the actual life difficulties that a student might face. As a student, the most miserable time of my life is during the final week of reviewing and preparing for exams, and we want to make this process easier. That's how we came up with the idea of TiCard.

What it does

It is a flashcard powered by AI. Unlike traditional flashcards, TiCard minimizes the amount of work we need to do as students. With TiCard, all you need to do is just enter the keyword of the subject that you need to study for, and TiCard will generate everything you need for you! No matter which program you are in, which subject you are interested in, math, science, literature, psychology and so on... TiCard has everything you need! Not only you need TiCard while preparing a exam, you can also use it when you are trying to learn about a whole new field of study. Keywords and explanations generated by TiCard can give you a quick idea of a subject that you are unfamiliar to.

How we built it

We used Kotlin, a JVM language, to write our backend codes. That is, call the Groq Cloud API, our AI engine, and process the returned data so it is in the form that can be displayed on the frontend. We took the advantage of the OpenAI API compatibility of the Groq API, so we can take the advantage system prompts to fine-tune the AI to allow us to get the output in the format we wanted. We then use HTML, CSS, and JavaScript for our frontend, so the data fetched can be displayed.

We also used MathJax to render LaTeX formulas in our frontend pages.

Challenges we ran into

Sometimes, the AI returns some data that doesn’t respect our data format and break our JSON parser, and we need to keep tuning the AI and handle related errors, so the program runs as expected.

Accomplishments that we're proud of

We have successfully tuned the AI so it can output in the format we want. We also successfully make our frontend pages renders the information we got (terms, definitions, formulas, etc.)

What's next for TiCard

We want to keep fine-tuning the AI so it can produce more meaningful terms, that suits the user the best. We also want to polish the frontend pages more so it looks better and provides a better user experience.