My-School-Buddy

Inspiration

My-School-Buddy is an application I built out of curiosity for AI and my passion for innovation. I was inspired by my passion to make the world a better place. I want to spend most of my life serving humanity and building amazing products that can further improve the standard of the average masses. I want to see a world of equality, a world where education is accessible to everyone around the world.

What it does

This is a fullstack app built with next.js.This application serves as a companion for school students and it helps aide their academic perfomance. It comprises of a variouse features such as a question and answer chatBot, a relative PDF search tool, a Video to text script translator, an autonomouse agent that helps student perform some task.

How we built it

The structure includes a Frontend and an API route(backend). The frontend layout contains the designs and functionalities including the logic of the application. The backend on the other hand is used to make calls to the LLM(large language model) using an LLM framework known as Langchain.

When users submit a prompt on the frontend, it get passed to the backend where the execution takes place using the OpenAI model to make the necessary calls and in some cases we store data in a vectore database for similiarity search. We then pass the response back to the frontend and the response get displayed to the user(students).

Challenges we ran into

There was'nt much challenges experienced except the limited resources that were available, by building the applications with javascript as opposed to python which is more common.

Accomplishments that we're proud of

This project did not only enhance my AI knowledge but it also did buttress my Software development skills, it made me think in more logical ways and indepth reasonings, It made me a Pragmatic developer.

What we learned

I learnt a lot while building this project, because I had to make tones of research and read lots of documentations. This project made me familiar with so many tools and implementations, i learnt how other LLM works, how to store data in other vectore database, how to query prompts, how to make calls and so many other things. And most importantly i learnt how do all these using Javascript, which is indeed an achievment as most documentation for AI development are made using python.

What's next for My-School-Buddy

I am so happy with the project, it opend my mind to various ideas and what could be acheived using AI as a helping tool for mankind. I hope to build many more applications like this one again. I look forward to taking this a step further by adding more features and turning it into a SAAS and also seek possible funding from investors inoder to scale it up.

Built With

• blockchain
• langchain
• next.js
• pinecone