Penfluence

You know AI is taking over when it can read your scribbles better than your classmates and professors! Struggling to find old notes before an exam? Well meet Penfluence, it has got your back.

Comment

Video Presentation

Here’s our demo video on Canvas! Couldn't upload to YouTube thanks to connectivity issues, but no way were we letting that stop us from sharing our hard work! Click on Link: link

Inspiration

One of our team members, Sadia works at tutoring center, and she often sees students showing up with notebooks full of messy, incomplete notes and assignments.

We know that not everyone can afford an iPad, and relying on scratch paper just isn’t practical. It gets lost, mixed up, or forgotten. That got us thinking, what if we could make it easier? Instead of letting lecture notes pile up or get tossed aside, why not just take a photo with your phone? It's a quick, simple way to keep everything organized and safe.

That’s how Penfluence came to life. A tool designed to make your handwritten notes work harder for you. With Penfluence, you can snap a picture of your notes, store them securely, and track your progress. It’s not about replacing paper, it’s about making it more useful.

What it does

Penfluence lets students snap a photo of their handwritten notes with their phones, turning them into digital format in an instant. The app then organizes everything, creates PDFs, and gives them a solid platform to study and review efficiently.

How we built it

For the frontend, we used Next.js with TypeScript to create a smooth, user-friendly interface. On the backend, we went with FastAPI to handle requests quickly and efficiently. We integrated OpenAI’s API to help with smart note recognition and analysis, and used OpenCV for processing images and manipulating documents. For storing and managing data, we set up MySQL with SQLAlchemy ORM, making it easy to query and handle everything behind the scenes.

Challenges we ran into

At first, we tried using OpenOCR, which a lot of our competitors rely on. But soon, we realized it just didn’t give us the level of accuracy we needed. So, even though switching back to OpenAI's API was a bit of a hassle, it ended up being totally worth it. With over 95% accuracy, we were able to set ourselves apart from other apps on the market.

Accomplishments that we're proud of

We’re really proud of how we managed to generate PDFs from extracted text and used OpenCV for processing images. Plus, finding a better alternative to OpenOCR really helped increase the accuracy and overall quality of our text extraction.

What we learned

This project taught us a lot, especially how to integrate FastAPI with MySQL to store and retrieve student data. We also got hands-on experience with OpenAI’s API for natural language processing and OpenCV for image processing, skills that really came in handy.

What's next for Penfluence

We’re working on building an AI tutor that activates as soon as a student takes a picture of their handwritten notes. The AI will immediately start tutoring based on the topic, helping students learn faster by answering questions and explaining the material in detail. We believe this will make it easier for students to grasp concepts, just by snapping a photo of their notes, they’ll have an AI tutor ready to guide them through it.

Instead of being overwhelmed by a whole notebook the night before an exam, students can break down their learning into small, manageable pieces every day. If they’re already reviewing and understanding the material as they take pictures to convert them into PDFs, it makes the whole process much easier. The future of Penfluence is all about making students more productive, one bite-sized chunk of learning at a time.

Built With

Share this project:

Updates