CleanSight: Data Quality & Preprocessing Application

CleanSight is a powerful tool that allows users to analyze, preprocess, and download high-quality datasets with just a few clicks. Built with FastAPI (backend) and Next.js (frontend), it simplifies the data cleaning process for analysts, developers, and data scientists.

Features

• Upload datasets in CSV, Excel formats.

• Paste datasets directly in the UI (CSV/JSON)

• Data Quality Report with:

  • Missing values
  • Column-wise stats
  • Outlier detection and Visualizations

• Preprocessing pipeline with:

  • Column cleaning
  • Imputation for missing values
  • Label Encoding for categorical features
  • Min-Max Scaling for numerical features

• Download: Data Quality Report (PDF), Preprocessed Dataset (CSV)

Tech Stack

• Backend: FastAPI, Pandas, scikit-learn, Seaborn, Matplotlib, FPDF
• Frontend: Next.js + Tailwind + Shadcn/UI
• Data Formats: CSV, Excel, JSON(pasted)

How It Works

Clone the repository:

git clone https://github.com/your-username/cleansight.git
cd cleansight

Set Up Locally By

Backend Setup

cd backend
python -m venv venv
source venv/bin/activate  # For Linux/Mac
venv\Scripts\activate     # For Windows
pip install -r requirements.txt
uvicorn main:app --reload

Backend runs at http://localhost:8000.

Frontend Setup

cd ../frontend
npm install
npm run dev

Frontend runs at http://localhost:5173.