JEE Study Planner (Prototype) 

A simple system to plan study tasks and see real progress during your JEE preparation. A personal tool designed around real JEE study workflows.  

Year: 2025
Status: Prototype
Type: Personal tool · UX-focused · Self-initiated project  

🧠 Overview

JEE Study Planner is a lightweight personal system I built to plan daily study tasks and track real progress during my JEE preparation. Instead of generic to-do apps, this prototype is designed around how JEE aspirants actually study — subjects, chapters, problem practice, and consistency.

The goal was not to build a “perfect app,” but a tool that feels motivating and realistic for long-term exam preparation.

.

🎯 Why I built this

Most study planners felt either too generic or too complex. I wanted something that:

• Matches real JEE study workflows

• Makes progress visible and motivating

• Stays simple enough to actually use every day

This project started as a personal need and became a UX-focused prototype.

🧩 Key Features

• Daily and topic-based study planning

• Clear visual progress tracking

• Minimal, distraction-free interface

• Designed specifically for JEE-style preparation

📚 What I learned

• Breaking a vague problem into clear features

• How small UX decisions affect motivation

• Why shipping an imperfect prototype is better than waiting for perfection

• Thinking like a user, not just a builder

🛠️ Tech & Approach

• Focus on logic, structure, and UX flow

• Built as a prototype to validate ideas quickly

• Prioritized usability over complexity
  
  

NoteFlow — Handwritten Notes Digitizer 

OCR-based system to convert handwritten notes into clean, exportable text.

Year: 2025
Status: Experimental Prototype
Type: Personal project · Systems + AI exploration

🧠 Overview

NoteFlow is an experimental notes application that converts handwritten notes into clean, readable digital text. The project explores the complete pipeline — from image upload and OCR to text processing and export — focusing on real-world challenges like handwriting accuracy, system reliability, and deployment.

This project was built primarily as a learning exercise to understand how real AI-powered tools are structured and deployed, rather than as a polished consumer product.

🎯 Why I built this

I wanted to explore:

• How handwritten OCR systems actually work in practice

• Why real-world AI products fail beyond just “model accuracy”

• The engineering challenges behind deployment, APIs, and reliability

Instead of stopping at “OCR works,” I pushed the project through hosting, backend services, and infrastructure decisions.

🧩 Key Features

• Upload handwritten notes as images

• Extract text using OCR pipelines

• Optional AI-based text cleanup and formatting

• Export output as PDF or DOCX

• Designed as a modular backend system

🛠️ Tech & Architecture

• Backend API for image processing

• OCR pipeline with preprocessing

• AI-assisted text post-processing (experimental)

• Cloud deployment and environment-based configuration

• Focus on fault tolerance and graceful failure

📚 What I learned

• OCR accuracy is as much about preprocessing as models

• Deployment and configuration are harder than writing code

• Why environment variables, APIs, and infra matter

• How to debug real production-style failures

• When to simplify features for stability