Introducing PDF Extraction Assistant: Extract Q&A from textbook PDFs. Ideal for educators and students, it transforms study materials into easy-to-use Q&A formats for efficient and effective learning.
Inspiration - I got this idea when I used to watch online videos of online tutors for my acadmic related subjects, they were preparing a lot of slides containing questions, and I thought If they are doing it all manually, what If I could make a tool that automates extracting Questions and answers from pdfs
What does it do - The users could upload pdfs that are textbooks or contains similar resources and my tool extracts the questions and answers from it. It also extracts the images as well.
How we built it - So considering that I have been working in web dev for around 2 years now, I picked NextJS, React for the frontend, django for the backend, Firebase storage for storing uploaded files, and also Firebase as the database. I first built out the backend as it was the most toughest part, that contains extracting text, images from pdf, detecting figures and stuff from images, constructing questions and answers from it, I used mistral for it. Then I went on to build the frontend, and finally deployed both of them.
Challenges we ran into - I was pretty new to PDF extraction, Image extraction and stuffs like these. So it took me some time to get comfortable in this new environment as this was also my first project working on django, I needed to try out some libraries to see what was the best and what worked best for the projects. A lot of times while working on this project, I did thought to give up and move on to another project. But I did knew that this project would take time and finally it would be worth it.
Accomplishments that we're proud of - I'm really proud of seeing the final product, everything working so seamlessly, even though there is a lot to improve, I have got the base mvp version. I'm pretty proud the I have successfully transformed a half cooked idea to a working project and looking forward of getting feedback from users and reaching out to people.
What we learned - I personally learnt a ton from this project, as I mentioned I have never worked in an environment especially, the backend. I became comfortable with using various python libraries for extracting text, images, etc. I learn about detecting objects from images, etc. I also got a good idea on how nextAuth worked as well as I'm using it for auth in the frontend. I'm really happy that I have learnt a lot by working on this project
What's next for PDF Q/A Extractor - I'm looking forward to get this product out to initial users, people who are in need of a tool like this, getting feedback from them, improving the product, growing it. I think it is such a helpful and useful tool for people working as educational instructors and also students
Log in or sign up for Devpost to join the conversation.