Inklyn

AI-powered document assistant that helps users understand documents, write emails, and answer questions based on uploaded files. Built with Streamlit and integrated with OpenRouter API for advanced AI responses.

Features

• Upload and process various document types: PDF, DOCX, TXT, CSV, Excel, PPTX, JSON, images with OCR, and more.
• Extract text, analyze data, and summarize content.
• Ask questions about uploaded documents.
• Compose and reply to professional emails using AI.
• Supports multiple AI models accessible via OpenRouter.
• User-friendly web interface with session management and chat history.

Installation

Prerequisites

• Python 3.8 or higher
• Tesseract OCR installed and configured (required for image OCR functionality). See: https://github.com/tesseract-ocr/tesseract
• An OpenRouter API key (free or paid, obtain from OpenRouter platform)

Setup

1. Clone the repository:

   • git clone https://github.com/manyaigdtuw/Inklyn
   • cd Inklyn

2. (Optional) Use the provided setup script to scaffold the environment and files: python setup_project.py

3. Install dependencies: pip install -r requirements.txt

4. Create a .env file in the root directory with your OpenRouter API key and app settings:

• OPENROUTER_API_KEY=your_openrouter_api_key_here
• OPENROUTER_BASE_URL=https://openrouter.ai/api/v1
• APP_NAME=AI Document Chatbot
• DEFAULT_MODEL=anthropic/claude-3-haiku

5. Run the application: streamlit run app.py

Usage

• Upload files via the Streamlit UI.
• The app automatically processes and extracts content from uploaded files.
• Chat with the AI assistant to get document explanations, summaries, or draft emails.
• Select different AI models from the sidebar as needed.

Project Structure

• app.py — Main Streamlit application.
• ai_engine.py — Interacts with OpenRouter API for AI-generated responses.
• document_processor.py — Processes uploaded documents (PDF, Word, images, spreadsheets, etc.) with OCR support.
• email_assistant.py — AI-powered email writing and replying helper.
• config.py — Configuration and environment variables.
• requirements.txt — Python dependencies.
• setup_project.py — Automated project setup script.

Supported File Types

• PDF, Word (.docx, .doc), Text, CSV, Excel, PowerPoint, JSON.
• Images: PNG, JPG, BMP, TIFF, WebP (with OCR extraction).
• Generic text for unrecognized file formats.

Technologies Used

• Streamlit — Web app framework.
• OpenRouter API — AI language model API.
• PyMuPDF — PDF parsing.
• python-docx — Word document parsing.
• Pandas — Data analysis and processing.
• Pillow — Image processing.
• PyTesseract & EasyOCR — Optical character recognition (OCR).
• python-pptx — PowerPoint file processing.

Contributing

Contributions are welcome! Please open issues or submit pull requests for improvements and fixes.

Acknowledgements

• Thanks to OpenRouter for providing AI APIs.
• Inspired by open-source projects for document processing and AI chatbots.