Spaces:
Sleeping
Sleeping
| <html lang="en"> | |
| <head> | |
| <meta charset="UTF-8"> | |
| <meta name="viewport" content="width=device-width, initial-scale=1.0"> | |
| <title>BMS AI Assistant - Technical Specifications</title> | |
| <style> | |
| :root { | |
| --primary: #D01010; | |
| --header-bg: #333; | |
| --light-bg: #f8f9fa; | |
| } | |
| body { | |
| font-family: 'Segoe UI', Arial, sans-serif; | |
| line-height: 1.6; | |
| color: #333; | |
| max-width: 1000px; | |
| margin: 0 auto; | |
| padding: 40px; | |
| } | |
| header { | |
| border-bottom: 4px solid var(--primary); | |
| margin-bottom: 40px; | |
| padding-bottom: 20px; | |
| } | |
| h1 { | |
| margin: 0; | |
| color: var(--header-bg); | |
| font-size: 2.5em; | |
| } | |
| h2 { | |
| color: var(--primary); | |
| margin-top: 40px; | |
| border-bottom: 1px solid #ddd; | |
| padding-bottom: 10px; | |
| } | |
| h3 { | |
| color: #555; | |
| margin-top: 25px; | |
| } | |
| table { | |
| width: 100%; | |
| border-collapse: collapse; | |
| margin: 20px 0; | |
| font-size: 0.95em; | |
| } | |
| th, | |
| td { | |
| padding: 12px 15px; | |
| border: 1px solid #ddd; | |
| text-align: left; | |
| } | |
| th { | |
| background-color: var(--header-bg); | |
| color: white; | |
| font-weight: 600; | |
| } | |
| tr:nth-child(even) { | |
| background-color: var(--light-bg); | |
| } | |
| .tech-grid { | |
| display: grid; | |
| grid-template-columns: repeat(2, 1fr); | |
| gap: 20px; | |
| } | |
| .tech-box { | |
| background: var(--light-bg); | |
| padding: 20px; | |
| border-left: 4px solid var(--primary); | |
| } | |
| .tech-box h4 { | |
| margin-top: 0; | |
| color: var(--primary); | |
| } | |
| ul { | |
| margin: 0; | |
| padding-left: 20px; | |
| } | |
| </style> | |
| </head> | |
| <body> | |
| <header> | |
| <h1>BMS AI Assistant</h1> | |
| <p><strong>System Technical Specifications</strong> | Version 1.0</p> | |
| </header> | |
| <h2>1. System Overview</h2> | |
| <p>The BMS AI Assistant is a specialized conversational interface designed for supply chain optimization. It | |
| integrates deterministic forecasting models with natural language processing to provide real-time decision | |
| support for inventory and procurement operations.</p> | |
| <h2>2. Technology Stack</h2> | |
| <div class="tech-grid"> | |
| <div class="tech-box"> | |
| <h4>Core Framework</h4> | |
| <ul> | |
| <li><strong>Language:</strong> Python 3.10</li> | |
| <li><strong>API Framework:</strong> FastAPI</li> | |
| <li><strong>Server:</strong> Uvicorn (ASGI)</li> | |
| </ul> | |
| </div> | |
| <div class="tech-box"> | |
| <h4>Data & Analytics</h4> | |
| <ul> | |
| <li><strong>Processing:</strong> Pandas, NumPy</li> | |
| <li><strong>Forecasting:</strong> Statsmodels (ARIMA)</li> | |
| <li><strong>Storage:</strong> Flat-file CSV (In-memory cache)</li> | |
| </ul> | |
| </div> | |
| <div class="tech-box"> | |
| <h4>Frontend</h4> | |
| <ul> | |
| <li><strong>Structure:</strong> HTML5</li> | |
| <li><strong>Styling:</strong> CSS3 (Custom Design System)</li> | |
| <li><strong>Logic:</strong> Vanilla JavaScript (ES6+)</li> | |
| </ul> | |
| </div> | |
| <div class="tech-box"> | |
| <h4>Deployment</h4> | |
| <ul> | |
| <li><strong>Containerization:</strong> Docker</li> | |
| <li><strong>Platform:</strong> Hugging Face Spaces</li> | |
| <li><strong>CI/CD:</strong> Git-based Workflow</li> | |
| </ul> | |
| </div> | |
| </div> | |
| <h2>3. Component Specifications</h2> | |
| <table> | |
| <thead> | |
| <tr> | |
| <th>Component</th> | |
| <th>Specification</th> | |
| <th>Function</th> | |
| </tr> | |
| </thead> | |
| <tbody> | |
| <tr> | |
| <td><strong>Forecasting Engine</strong></td> | |
| <td>ARIMA (1,1,1) Model</td> | |
| <td>Generates 30-day demand projections based on historical time-series data.</td> | |
| </tr> | |
| <tr> | |
| <td><strong>Intent Parser</strong></td> | |
| <td>Regex / Pattern Matching</td> | |
| <td>Identifies user intent (Forecast, Inventory, Supplier) with <50ms latency.</td> | |
| </tr> | |
| <tr> | |
| <td><strong>LLM Integration</strong></td> | |
| <td>TinyLlama 1.1B (GGUF)</td> | |
| <td>Handles unstructured general queries and conversational context.</td> | |
| </tr> | |
| <tr> | |
| <td><strong>PDF Generator</strong></td> | |
| <td>FPDF Library</td> | |
| <td>Dynamically compiles chat transcripts and data tables into downloadable reports.</td> | |
| </tr> | |
| </tbody> | |
| </table> | |
| <h2>4. Performance Metrics</h2> | |
| <ul> | |
| <li><strong>API Latency:</strong> < 100ms (Standard Queries), < 2s (Forecasting)</li> | |
| <li><strong>Concurrent Sessions:</strong> Scalable via Docker replicas (Default: 2 vCPUs)</li> | |
| <li><strong>Uptime:</strong> 99.9% (Hugging Face Infrastructure)</li> | |
| </ul> | |
| </body> | |
| </html> |