"A high-performance, AI-driven platform designed to revolutionize document interaction. By leveraging the Gemini API and a robust FastAPI backend, Doc Assistant enables users to distill complex PDF/TXT data into actionable insights through automated summarization and semantic Q&A."
Doc Assistant is a professional-grade analysis tool developed by Emine Uğurlu. It addresses the challenge of information overload by providing a scalable environment for instant document parsing, keyword search, and intelligent dialogue with static files.
This project showcases advanced backend orchestration and AI service integration:
- Gemini AI Integration: Implementation of sophisticated prompt engineering within
ai_service.pyto deliver high-context summaries and precise Q&A. - Asynchronous Backend Architecture: Utilizing FastAPI to manage non-blocking I/O operations for seamless file uploads and real-time AI processing.
- Document Parsing Engine: Robust text extraction and chunking logic for PDF and TXT formats handled by a dedicated
file_processor.py. - Relational Data Management: Structured storage of document metadata and user interactions using SQLite with efficient CRUD operations.
- Scalable Routing Layer: Modular API design with separate routers for AI chat, search, and document management.
- 🧠 Semantic Q&A: Ask complex questions and receive context-aware answers directly from your documents.
- 📝 Automated Summarization: Instantly generate executive summaries for long-form PDF and TXT files.
- 🔍 Precision Search: Deep-file keyword search engine to locate critical information across your library.
- 🗂️ Document Management: Fully interactive dashboard to upload, view, and manage your analyzed documents.
- Backend: FastAPI, Python, Pydantic.
- AI Engine: Google Gemini API.
- Database: SQLite.
- Frontend: HTML5, CSS3, JavaScript (Vanilla).
- Clone the Repository:
git clone https://github.com/emineugurlu/doc-assistant
cd doc-assistant2.Environment Configuration: Create a .env file and add your GEMINI_API_KEY.
3.Install Dependencies:
pip install -r requirements.txt4.Launch the Server:
uvicorn main:app --reloadDeveloped by Emine Uğurlu - Computer Engineer Empowering document intelligence through advanced engineering.