Offline AI Document Analysis Tool

I developed this software in 2024 during my internship at the United Nations Headquarters in New York. It was the first project I worked on in the field of artificial intelligence, and I had to develop the software within two months and obtain approval from the review board. The software and its architecture were officially approved by the UN Secretariat and integrated into the United Nations inventory as an official application. Despite the class crises I faced during my time there—as well as the segregation efforts by certain financial hegemonic stakeholders of the West, their awkward identity crises, and the pressures of radicalized far-right tendencies—it was a source of pride to persevere through these challenges, complete the project, and successfully navigate the approval process.


GitHub GitHub Repository

This tools functions as an Offline AI Document Analysis Tool with Natural Language Processing, search, and data visualization functionalities. The tool comprises six tabs to address specific needs within document management, analysis, and visualization. Supported file formats for document analysis includes PDF, DOCX and TXT.

The main purpose of developing this tool was to analyze documents via AI in an offline setting, allowing it to function without an internet connection. You are welcome to edit and improve this tool.

Installation & Running Instructions

  1. Install the required packages:
pip install -r requirements.txt
  1. Download the English language model for SpaCy:
python -m spacy download en_core_web_lg
  1. Run the Application:
python main.py

Deployment Instructions

To deploy the application as a standalone executable, you can use PyInstaller with the following command:

pyinstaller -D -w --collect-all pyvis --additional-hooks-dir=. main.py

Preview

Note: The open source documents analyzed in the preview were retrieved from https://main.un.org/securitycouncil/en/sanctions/1267/aq_sanctions_list/summaries in December 2023.

imageimageimageimageimageimageimageimageimage