Data Preprocessing AI Agent

An interactive Streamlit-based AI tool that automates data preprocessing and exploratory data analysis (EDA) for CSV datasets using LLM-powered explanations.

Upload your dataset, select your machine learning goal (e.g., classification, regression, clustering, or text analysis), and receive a step-by-step, explainable preprocessing pipeline. The tool also provides cleaned data and a downloadable PDF report.

Features

• Automates EDA: Identifies dataset shape, data types, missing values, outliers, and correlations

• Adapts to ML goals: Preprocessing steps change based on your selected task (classification, regression, etc.)

• Explains every step: Uses a language model to describe preprocessing decisions in simple terms

• Visualizes data: Generates histograms, boxplots, heatmaps, and class balance charts

• Supports key preprocessing tasks:

  • Missing value imputation
  • Outlier detection and removal
  • Feature scaling (MinMax, Standard)
  • Encoding (Label, OneHot)
  • Power transformation
  • Low-variance feature removal
  • Text normalization

• Provides downloadable results:

  • Cleaned dataset as CSV
  • EDA and preprocessing summary as PDF

---

Tech Stack

• Python 3.10+
• Streamlit
• Pandas, NumPy
• Seaborn, Matplotlib
• Scikit-learn, SciPy
• FPDF or ReportLab for PDF generation
• Ollama LLM Model

---

How to Use

• Clone the repository:

  git clone https://github.com/abhay-2108/Data-Preprocessing-AI-Agent.git
  cd Data-Preprocessing-AI-Agent

• Install dependencies:

  pip install -r requirements.txt

• Run the app:

  streamlit run data_preprocessor_tool.py

• Upload your CSV, choose your ML task, and view the results with explanations and downloads

---

License

• This project is licensed under the MIT License. See the LICENSE file for more details.

---