-
-
The view of application when opened at first
-
It shows a error message when the server is busy and does not collapse
-
Sample image of handwritten notes
-
Sample image of handwritten notes
-
answer for question
-
question
Inspiration
As a first-year B.Tech student, I mostly study from handwritten classroom notes and a few shared PDFs. Before exams, I spend a lot of time rewriting notes, creating revision points and framing answers for practice. I noticed that most AI tools behave like general chatbots and do not work directly with a student’s own study material. I wanted to build something that could actually help during real exam preparation using the notes I already have. This personal study experience inspired me to build FocusStudy AI.
What it does
FocusStudy AI is an exam-focused study companion. It allows students to upload typed PDF notes or images of handwritten notes and ask questions in three focused modes: Explain, Exam Ready, and Revision. The answers are generated strictly using the uploaded notes, helping students understand concepts, prepare structured exam answers, and revise quickly before tests.
How we built it
I built the project as a web application using Python and Flask for the backend and HTML/CSS for the frontend. Typed PDF text is extracted using PyPDF2. For handwritten notes, I use Gemini’s multimodal capability to directly read and understand text from uploaded images. Gemini’s text generation is then used to produce answers according to the selected study mode. The application is deployed publicly using Railway and uses the Google Gen AI SDK to interact with Gemini.
Challenges we ran into
The biggest challenge was handling handwritten notes. I initially tried local OCR tools, but they required heavy system dependencies and created multiple installation and deployment issues. This made the application unreliable. I solved this problem by switching to Gemini’s multimodal processing, which removed the need for local OCR and significantly improved stability.
Accomplishments that we're proud of
I am proud that I was able to build and deploy a complete working application as a solo first-year student. The project supports both typed and handwritten notes and provides exam-oriented answer modes instead of generic chat responses. Successfully deploying a real AI application and making it accessible on both desktop and mobile browsers is one of my biggest achievements in this project.
What we learned
Through this project, I learned how real AI systems are designed and deployed end-to-end. I learned about file handling, prompt design for different learning goals, multimodal AI integration, and cloud deployment. I also learned the importance of choosing reliable and simple engineering solutions instead of complex tools that are difficult to maintain.
What's next for FocusStudy AI
In the future, I would like to add session-based study tracking, allow students to save important answers, and support multi-page handwritten uploads in a smoother way. I also plan to improve accessibility features and further optimize prompts for different academic subjects to make FocusStudy AI more personalized and effective for students.
Log in or sign up for Devpost to join the conversation.