Interactive AI-powered report generator with visualizations and export to PDF/PPTX/HTML. Built with Streamlit and OpenAI.
AutoReport-AI is an interactive Streamlit application that automatically generates insightful reports from CSV or Excel datasets using artificial intelligence. Key features include:
- 📊 Data summarization using GPT-4 (OpenAI) to describe the dataset's structure, columns, statistics, and more.
- 📈 Automatic visualizations for all numeric variables with histograms and boxplots.
- 🖼️ AI-generated image based on the data description using DALL·E.
- 🧠 Python code generation with GPT-4 to build custom visualizations.
- 📤 Export options: create reports in PDF, PowerPoint, and HTML formats.
Ideal for data analysts, data scientists, and non-technical users who need to quickly turn data into beautiful, shareable reports — without writing code manually.
AutoReport-AI is an interactive Streamlit application that automates the generation of data reports using artificial intelligence. It analyzes datasets, creates visualizations, and exports complete reports in PDF, PowerPoint, and HTML formats — all without writing code.
- Upload
.csvor.xlsxdatasets. - Preview the dataset in a friendly Streamlit interface.
- Uses GPT-4 (OpenAI) to generate a descriptive summary:
- Describes each column.
- Explains ranges, measures, and distributions.
- Available in English or Spanish.
- Creates:
- Histograms and boxplots for all numeric columns (via Plotly).
- Additional matplotlib plots generated dynamically by GPT-4.
- All plots are rendered live and saved as
.png.
- Generates a unique image that visually represents the dataset based on its description using DALL·E.
- Export the full report in:
- PowerPoint (.pptx)
- HTML
- Each export includes:
- Summary, visualizations, code-generated charts, and DALL·E image.
- Allows uploading of additional
.txtfiles (description.txt,summary.txt) to manually provide summaries. - All files (uploaded data, generated images, reports, plots) are saved in the same folder where the app is executed, making them easily accessible after the session ends.
- Supports multiple report generation workflows in the same run.
- Clone the repository:
git clone https://github.com/your-username/AutoReport-AI.git
cd AutoReport-AI- Create a virtual environment and activate it (optional but recommended):
python -m venv venv
source venv/bin/activate # On Windows: venv\Scripts\activate- Install the dependencies:
pip install -r requirements.txt- Set your OpenAI API key:
# Inside the code (temporarily, or replace with environment variable):
openai.api_key = "your-openai-api-key"streamlit run main.pyThen open the app in your browser (usually http://localhost:8501).
Make sure you have the following Python libraries:
streamlitpandasplotlyfpdfmatplotlibopenaiPillowpython-pptxrequests
You can install them all via:
pip install -r requirements.txtComing soon...
MIT License ©bermejo4