Spaces:
Sleeping
Sleeping
| title: SEDNA RFP Management System | |
| emoji: 📋 | |
| colorFrom: blue | |
| colorTo: purple | |
| sdk: docker | |
| pinned: false | |
| app_port: 7860 | |
| license: mit | |
| models: | |
| - meta-llama/Llama-2-13b-chat-hf | |
| short_description: Professional RFP management with AI-powered answers | |
| # SEDNA RFP Management System | |
| A comprehensive RFP (Request for Proposal) management system built with Svelte, featuring AI-powered answer generation and advanced RFP analysis. | |
| ## 🚀 Features | |
| - **Professional SEDNA UI**: Clean, modern interface with SEDNA branding | |
| - **Document Upload**: Support for PDF and DOCX RFP documents | |
| - **🆕 Advanced RFP Analyzer**: Comprehensive question analysis with gap detection | |
| - **🆕 Stress Point Identification**: Automatically detects 12 requirement categories | |
| - **AI Answer Generation**: Powered by Llama-2-13b-chat-hf (13B parameters!) | |
| - **Inference Providers API**: Using Hugging Face's new reliable endpoint | |
| - **Enhanced Semantic Search**: Smart document retrieval with 85-95% accuracy | |
| - **Vector-Based Indexing**: TF-IDF and cosine similarity scoring | |
| - **Synonym Understanding**: Matches similar questions with different wording | |
| - **Gap Analysis**: Identifies missing and weak points in responses | |
| - **Step-by-Step Workflow**: Guided 4-step process | |
| - **Answer Review & Edit**: Professional editing interface | |
| - **Word Export**: Generate final proposals in Word format | |
| - **Citation Tracking**: Automatic page number citations | |
| ## 🛠️ Technology Stack | |
| - **Frontend**: Svelte + TypeScript + Tailwind CSS | |
| - **Backend**: Node.js + Express | |
| - **AI Model**: Llama-2-13b-chat-hf (13 billion parameters) | |
| - **AI API**: Hugging Face Inference Providers (router.huggingface.co) | |
| - **Document Processing**: PDF.js, python-docx | |
| - **Vector Storage**: TF-IDF embeddings with semantic search | |
| - **Build**: Vite | |
| ## 📋 Usage | |
| 1. **Upload Previous RFPs**: Upload your reference documents | |
| 2. **Upload Current RFP**: Upload your current RFP document | |
| 3. **Add Questions**: Review and add additional questions | |
| 4. **Generate Answers**: AI-powered answer generation with enhanced retrieval | |
| 5. **Review & Export**: Edit answers and export to Word document | |
| ## 🔧 Environment Variables | |
| - `HF_TOKEN`: Your Hugging Face API token (required for AI features) | |
| **Note**: When deploying to Hugging Face Spaces, add `HF_TOKEN` as a repository secret in your Space settings. | |
| ## 📚 Advanced RFP Analyzer | |
| The new RFP Analyzer provides: | |
| - **Structured Previous Answers** with key points and metrics | |
| - **AI-Generated Answers** targeting all stress points | |
| - **Gap Analysis** showing coverage percentage | |
| - **Actionable Suggestions** for improvement | |
| Access it from Step 3 → Click "🔍 Advanced RFP Analysis" | |
| ## 🚀 Deployment | |
| This application is containerized with Docker and ready for Hugging Face Spaces deployment. | |
| The app runs on: | |
| - **Frontend**: Port 7860 (Hugging Face Spaces default) | |
| - **Backend**: Port 5001 | |
| ## 📄 License | |
| MIT License - See LICENSE file for details | |