LearnDer

Teacher's example dashboard
Example of student AI chatbot interface

Inspiration

We looked back to when we were studying in high school, and we realized how difficult it is to deal with handed-out textbooks that happen to be too convoluted to study efficiently. We've also realized how difficult it is to learn everything about our class from ChatGPT since it requires a correct way to ask for one topic, and given that it can be answered wrong makes ChatGPT not much of an ideal way to learn. Which is why we chose to create this web application.

What it does

LearnDer is a web application that uses LLM to answer students with their exact class textbook information. The textbook is required to be added by their teachers as we provide teachers with their class student's information and insights, potentially offering what teachers need to improve their teachings.

How we built it

We used Bolt.new to generate frontends and then adjusted it to match our UX/UI design with Bolt.new and self-adjustments by ourselves. Backends are implemented into the frontend files with Bolt.new and then ChatGPT for the most of time.

Challenges we ran into

Each of our team members does not have much time to work so we only finish up to 60-75% of the desired tasks and features.
Our backend team members happen to be too busy, leaving only non-backend members to work on this for the most of time. It is also a reason why backends are implemented into the front end directly.
Bolt.new happens to be too heavily limited with daily tokens, making it not so ideal choice for hackathon participants to use. Regardless of being a very innovative and helpful service to both front and back-end developers, having to continue working on the generated work after running out of tokens is a huge hassle for our workflow. Not to mention subscribing to the pro version for the short of tokens stands out to be too expensive for such a short-time event like hackathons.

Accomplishments that we're proud of

We've accomplished core functions for this work even when we were heavily assigned with work and burden at the same time when we developed this web application.

What we learned

We learned the great use of Bolt.new for easing frontends and backends work that reduces time to develop every possible thing.

What's next for LearnDer

In our free time, we will be developing the missing percentage of our goal. Such as a teacher's dashboard with AI-assisted insight and data. Possibly also add a chatbot function to the teachers' side.
Once we are sure that this work is finished, we would like to pitch our plans to schools across Thailand, hopefully, this would help Thai schools to be much more in reach to AI worlds and ensure that they are not left behind.