{"id":29503820,"url":"https://github.com/ss-369/intellidoc-ai","last_synced_at":"2025-07-15T23:06:22.024Z","repository":{"id":304513307,"uuid":"1018854008","full_name":"ss-369/IntelliDoc-AI","owner":"ss-369","description":"A production-ready RAG-based Document Q\u0026A system using Streamlit and FastAPI. Combines ChromaDB for semantic retrieval and Gemini 2.5 Flash for accurate, grounded answer generation. Includes TruLens-powered evaluation for measuring answer quality, relevance, and source attribution.       ","archived":false,"fork":false,"pushed_at":"2025-07-13T14:43:20.000Z","size":1441,"stargazers_count":0,"open_issues_count":0,"forks_count":0,"subscribers_count":0,"default_branch":"main","last_synced_at":"2025-07-13T16:26:58.414Z","etag":null,"topics":["chromadb","fastapi","gemini-api","question-answering","rag","trulens","vector-database"],"latest_commit_sha":null,"homepage":"","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/ss-369.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null}},"created_at":"2025-07-13T07:30:41.000Z","updated_at":"2025-07-13T14:45:24.000Z","dependencies_parsed_at":"2025-07-13T16:27:05.104Z","dependency_job_id":"1b28e349-d9a6-4482-b4f2-19051671714c","html_url":"https://github.com/ss-369/IntelliDoc-AI","commit_stats":null,"previous_names":["ss-369/intellidoc-ai"],"tags_count":null,"template":false,"template_full_name":null,"purl":"pkg:github/ss-369/IntelliDoc-AI","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/ss-369%2FIntelliDoc-AI","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/ss-369%2FIntelliDoc-AI/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/ss-369%2FIntelliDoc-AI/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/ss-369%2FIntelliDoc-AI/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/ss-369","download_url":"https://codeload.github.com/ss-369/IntelliDoc-AI/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/ss-369%2FIntelliDoc-AI/sbom","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":265466216,"owners_count":23770507,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["chromadb","fastapi","gemini-api","question-answering","rag","trulens","vector-database"],"created_at":"2025-07-15T23:01:29.817Z","updated_at":"2025-07-15T23:06:22.013Z","avatar_url":"https://github.com/ss-369.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"# 🤖 Intelligent Document Q\u0026A System\n\nA production-ready Retrieval-Augmented Generation (RAG) system for intelligent document question-answering with comprehensive evaluation capabilities.\n\n## 🏗️ Architecture\n\n![System Architecture](screenshots/Mermaid%20Chart%20-%20Create%20complex,%20visual%20diagrams%20with%20text.%20A%20smarter%20way%20of%20creating%20diagrams.-2025-07-13-143054.png)\n*Complete RAG system architecture showing data flow from document upload to answer generation*\n\n![Sequence Diagram](screenshots/sequencedia.png)\n*Detailed sequence diagram showing component interactions and data flow*\n\n- **Frontend**: Streamlit web interface with real-time configuration\n- **Backend**: Dual-mode backend with:\n  - **Streamlit** for interactive user experience\n  - **FastAPI** for RESTful API access (`/upload`, `/ask`, `/evaluate`)\n- **Vector Database**: ChromaDB for persistent document embeddings\n- **LLM**: Google Gemini 2.5 Flash for answer generation\n- **Evaluation**: TruLens framework for RAG quality assessment\n\n\n## 🚀 Features\n\n![Application Interface](screenshots/interface.png)\n*Main application interface with document upload, Q\u0026A, and evaluation tabs*\n\n### Core Capabilities\n\n- **Multi-format Document Processing**: PDF, TXT with intelligent chunking\n- **Advanced Retrieval**: Semantic search with contextual compression\n- **AI-Powered Q\u0026A**: Context-aware answer generation with source attribution\n- **Quality Evaluation**: Comprehensive metrics for answer relevance and groundedness\n- **Interactive Interface**: Real-time parameter adjustment and conversation history\n\n### Technical Highlights\n\n- **Production-Ready**: Environment-based configuration and comprehensive logging\n- **Scalable Architecture**: Modular design supporting multiple LLM and embedding providers\n- **Persistent Storage**: ChromaDB with automatic document versioning\n- **Performance Monitoring**: Real-time evaluation metrics and statistics\n- **Source Attribution**: Document-level source tracking and citation\n\n![Document Processing](screenshots/doc_processing.png)\n*Document processing workflow showing chunking and vector storage*\n\n![Source Attribution](screenshots/doc_sources.png)\n*Source attribution and document tracking in Q\u0026A responses*\n\n## 📊 System Capabilities\n\n- **Processing**: Efficient document chunking and vector storage with ChromaDB\n- **Retrieval**: Fast semantic search with configurable similarity thresholds\n- **Scalability**: Modular architecture designed for high-volume document processing\n- **Evaluation**: Comprehensive quality assessment using TruLens framework\n- **Reliability**: Production-ready error handling and session management\n\n## 🛠️ Technology Stack\n\n| Component | Technology | Purpose |\n|-----------|------------|---------|\n| **Frontend** | Streamlit | Interactive web interface |\n| **LLM** | Google Gemini 2.5 Flash | Answer generation |\n| **Vector DB** | ChromaDB | Document embeddings storage |\n| **Processing** | LangChain, PyPDF2 | Document processing pipeline |\n| **Evaluation** | TruLens | RAG quality assessment |\n| **Configuration** | Environment Variables | Secure settings management |\n\n## 🚀 Quick Start\n\n1. **Install Dependencies**:\n\n   ```bash\n   pip install -r requirements.txt\n   ```\n\n2. **Set API Keys**:\n\n   ```bash\n   export GEMINI_API_KEY=\"your_gemini_api_key\"\n   ```\n\n3. **Run Application**:\n\n   ```bash\n   streamlit run app.py --server.port 5000\n   ```\n\n4. **Access Interface**: \u003chttp://localhost:5000\u003e\n\n## 📸 Demo Screenshots\n\n### Main Application Interface\n\n![Application Interface](screenshots/interface.png)\n\n### Document Processing Workflow\n\n![Document Processing](screenshots/doc_processing.png)\n\n### Source Attribution \u0026 Q\u0026A\n\n![Source Attribution](screenshots/doc_sources.png)\n\n### TruLens Evaluation Dashboard\n\n![TruLens Evaluation](screenshots/trulens.png)\n\n## 📁 Project Structure\n\n```\nIntelligentDocumentQA/\n├── app.py                     # Main Streamlit application\n├── utils/                     # Core utilities\n│   ├── config.py             # Configuration management\n│   ├── document_processor.py # Document loading and chunking\n│   ├── vector_store.py       # Vector database operations\n│   ├── retriever.py          # Advanced retrieval logic\n│   ├── qa_chain.py           # Question answering with LLM\n│   └── evaluation.py         # RAG evaluation framework\n├── .streamlit/config.toml    # Streamlit configuration\n└── chroma_db/                # Persistent vector database\n```\n\n## 🎯 Use Cases\n\n- **Enterprise Knowledge Management**: Internal document Q\u0026A systems\n- **Customer Support**: Automated FAQ and documentation assistance\n- **Research \u0026 Analysis**: Academic paper and report analysis\n- **Legal Document Review**: Contract and compliance document processing\n- **Educational Tools**: Textbook and course material Q\u0026A systems\n\n## 🔮 Roadmap\n\n- [ ] Multi-LLM support (OpenAI, Claude, local models)\n- [ ] REST API endpoints\n- [ ] Advanced caching and performance optimization\n- [ ] Multi-language support\n- [ ] Enterprise authentication integration\n\n## 📈 Performance \u0026 Evaluation\n\n![TruLens Evaluation](screenshots/trulens.png)\n*TruLens evaluation dashboard showing comprehensive RAG quality metrics*\n\nThe system includes comprehensive evaluation using TruLens framework:\n\n- **Answer Relevance**: Measures how well answers address questions\n- **Context Relevance**: Evaluates retrieval quality and relevance\n- **Groundedness**: Assesses answer factual accuracy against sources\n- **Real-time Metrics**: Live performance monitoring and statistics\n\n---\n\n*Built with ❤️ using modern AI/ML best practices and production-ready architecture.*\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fss-369%2Fintellidoc-ai","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fss-369%2Fintellidoc-ai","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fss-369%2Fintellidoc-ai/lists"}