Highlightr

Highlightr is an AI-powered document semantic markup specialist designed to process and highlight specific categories of information within various document formats. Using Google Gemini models, it identifies key points, problems, statistics, and more, providing a color-coded visual output and downloadable Word documents.

Features

• Multi-format Support: Upload and process .txt, .docx, and .pdf files.
• Semantic Tagging: Automatically identifies and tags information based on categories like Key Points, Problem Areas, Statistics, Opportunities, Actions, and KPIs.
• Custom Instructions: Provide custom highlighting logic via the AI prompt.
• Interactive Preview: Filter and view highlighted text directly in the browser.
• Export to DOCX: Download processed documents with original formatting and color-coded highlights.
• Token Management: Integrated token counter to manage API limits.

Tech Stack

• Backend: Python, Flask
• AI Engine: Google GenAI SDK (Gemini)
• Document Processing: python-docx, PyPDF2, BeautifulSoup4 (for tag parsing)
• Frontend: HTML5, CSS3, Vanilla JavaScript, Lucide Icons

Installation

1. Install dependencies:

   pip install flask python-docx google-genai beautifulsoup4 PyPDF2

2. Set up your Gemini API Key in the application settings.

3. Run the application:

   python app.py

Usage

1. Upload documents to the drop zone.
2. Select the highlighting options or provide a custom request.
3. Enter your Google AI Studio API key.
4. Click "Process files" to generate highlighted versions.
5. Use the preview eye icon to view results or download the generated .docx files.