Learn Smart GPT

Sick of expensive study coaches or wasting time on study plans? Our AI solution generates personalized study plans for any topic! Quench your intellectual curiosity and accelerate your learning today.

Comment

Inspiration

At the heart of Learn Smart GPT is our belief that everyone deserves access to a quality mentor that can guide their learning journey. However, we recognize that traditional study coaches can be expensive, with costs reaching up to $250 per session, making it out of reach for many people. Additionally, learning new topics and creating a personalized study plan can be time-consuming. That's why we were inspired to create an AI-powered tutor and study coach that is knowledgeable in any topic and can provide users with personalized suggestions for their study path. We envisioned a solution that would eliminate the barriers of cost and time, making it accessible to anyone who wants to learn and grow. With Learn Smart GPT, we hope to empower users to achieve their learning goals and realize their full potential.

What it does

Our solution starts by prompting the user about the topics they want to learn. From there, our ChatGPT 3.5 powered solution generates a comprehensive overview of the chosen topic, coupled with real-life applications, and suggests prerequisite knowledge to explore. With just a few clicks, our users can delve into the details of any topic they're interested in, making learning both accessible and engaging. Finally, our solution generates a detailed study plan tailored to the user, taking into account their unique preferences and learning style. We aim to make the learning journey as seamless and efficient as possible, enabling users to achieve their learning goals with ease.

How we built it

  • To build the front-end, we chose React for its ease of learning and the opportunity for workshops during the hackathon.
  • We accelerated our styling process with TailwindCSS, which allowed us to create a polished and visually appealing interface.
  • For the back-end, we opted for the simplicity of Express.js. The back-end receives user prompts from the front-end, and then sends a HTTP POST request to the OpenAI API with additional context and instructions. This enables ChatGPT to act as a knowledgeable mentor, providing users with personalized guidance and suggestions for their study plan.

Challenges we ran into

  • Learning new skills on the spot during our first hackathon was a challenge, as most of us had little experience with front-end development, back-end development, and Git collaboration
  • Integrating the OpenAI API into our product proved to be a challenge, especially during debugging Streaming GPT's response chunk by chunk was not natively supported in the npm openai package, and implementing the workaround with Server Side Events or Web Sockets was unfeasible due to time constraints
  • Bringing our complex Figma design to reality was a challenge that we had to compromise on due to time constraints, but we're proud of the final product's UI/UX

Accomplishments that we're proud of

  • A working product with a pleasant UI/UX
  • Integration with external API
  • First hackathon for most of us, and our first project powered by AI

What we learned

  • React front-end development
  • Back-end development with Express.js
  • Figma design
  • Integrating external API into our product
  • Prompt engineering

What's next for Learn Smart GPT

  • Database and platform where students and teachers can share their plan
  • Implement peer review system where teachers can rate someone's learning path/study plan
  • Notification feature urging students to learn
  • When students mark their study plan as done, our product will test them regarding the topics and mark their understanding (test and marking done by ChatGPT)
  • When students make mistakes, ChatGPT analyses the mistake, create a new study plan based on those mistake, so students can revise their weak points

Check our repo, design, and slides

Built With

Share this project:

Updates