# AInSight - Enterprise AI System A full-fledged Enterprise AI System with privacy-first on-device processing. ## 🌟 Overview **AInSight** is a comprehensive enterprise AI platform that combines: - **Web Dashboard** for admin management, analytics, and team collaboration - **Mobile App** with RunAnywhere SDK for offline AI processing - **Privacy-First Architecture** - all sensitive data processed on-device ## 🎯 Key Features ### 🧠 On-Device AI Assistant (Mobile) - ✅ Smart document summaries - ✅ Professional email drafting - ✅ Code assistance - ✅ Task notes & voice memos - ✅ Runs 100% offline using RunAnywhere SDK ### 📊 Secure Admin Dashboard (Web) - ✅ Employee access management - ✅ Usage analytics & productivity trends - ✅ Company-wide announcements - ✅ Feature access control ### 🔐 Privacy & Compliance - ✅ All AI inference runs locally on devices - ✅ Only encrypted metadata sent to server - ✅ GDPR & HIPAA compliant architecture - ✅ Perfect for banks, hospitals, and startups ### 💬 AI Workspace Chat - ✅ Encrypted P2P team communication - ✅ Private AI-powered chat summaries (on-device) - ✅ No server-side message storage ### 🎤 Voice-to-Note + Task Extraction - ✅ Offline voice transcription - ✅ AI-powered task extraction - ✅ Auto-sync to dashboard ## 📁 Project Structure ``` AInSight/ ├── backend/ # Python API Server │ ├── app/ │ │ ├── models/ # Database models │ │ ├── routes/ # API endpoints │ │ ├── utils/ # Helpers & utilities │ │ └── config.py # Configuration │ ├── requirements.txt │ └── run.py │ ├── frontend/ # Admin Dashboard │ ├── index.html │ ├── css/ │ ├── js/ │ └── assets/ │ ├── mobile/ # Mobile App (React Native) │ ├── src/ │ │ ├── components/ │ │ ├── screens/ │ │ ├── services/ # RunAnywhere SDK integration │ │ └── utils/ │ ├── package.json │ └── README.md │ └── docs/ # Documentation ├── API.md ├── SETUP.md └── ARCHITECTURE.md ``` ## 🚀 Quick Start ### Prerequisites - Python 3.8+ - Node.js 16+ - React Native environment (for mobile) - RunAnywhere SDK account ### Backend Setup ```bash cd backend pip install -r requirements.txt python run.py ``` ### Frontend Setup ```bash cd frontend # Open index.html in browser or serve with: python -m http.server 8080 ``` ### Mobile Setup ```bash cd mobile npm install npm run android # or npm run ios ``` ## 🏗️ Architecture ``` ┌─────────────────┐ ┌──────────────────┐ │ Mobile App │◄───────►│ Web Backend │ │ (RunAnywhere) │ HTTPS │ (REST API) │ └─────────────────┘ └──────────────────┘ │ │ │ │ ▼ ▼ ┌─────────────────┐ ┌──────────────────┐ │ On-Device AI │ │ Admin Dashboard │ │ (Private) │ │ (Analytics) │ └─────────────────┘ └──────────────────┘ ``` ## 🔒 Security Features - **End-to-End Encryption** for all communications - **On-Device Processing** via RunAnywhere SDK - **Zero-Knowledge Architecture** - server never sees raw data - **Token-Based Authentication** (JWT) - **Role-Based Access Control** (RBAC) ## 📱 Mobile Features 1. **Document Summarizer** - Summarize PDFs, docs, meeting notes 2. **Email Assistant** - Draft professional emails with AI 3. **Code Helper** - Get code suggestions & debugging help 4. **Task Manager** - AI-powered task prioritization 5. **Voice Memos** - Speech-to-text with task extraction ## 🖥️ Dashboard Features 1. **User Management** - Add/remove employees, set permissions 2. **Analytics** - Usage stats, productivity insights 3. **Announcements** - Broadcast messages to mobile app 4. **Access Control** - Enable/disable features per user 5. **Audit Logs** - Track system usage (privacy-safe) ## 🛠️ Tech Stack ### Backend - **Framework**: Flask/FastAPI - **Database**: PostgreSQL/SQLite - **Auth**: JWT tokens - **Encryption**: AES-256, RSA ### Frontend - **UI**: HTML5, CSS3, JavaScript (Vanilla/React) - **Charts**: Chart.js / D3.js - **API Client**: Axios ### Mobile - **Framework**: React Native / Flutter - **AI SDK**: RunAnywhere SDK - **Storage**: SQLite (local) - **Networking**: Axios ## 📝 API Endpoints - `POST /api/auth/login` - Admin login - `GET /api/users` - List employees - `POST /api/users` - Add employee - `GET /api/analytics` - Usage statistics - `POST /api/announcements` - Send announcement - `GET /api/tasks/:userId` - Sync user tasks ## 🌐 Local AI Integration Options AInSight supports multiple local AI SDKs for privacy-first, offline AI processing: ### Option 1: **Transformers.js** (Recommended for Web) ⭐ ```javascript import { pipeline } from '@xenova/transformers'; // Load model (cached for offline use) const summarizer = await pipeline('summarization', 'Xenova/distilbart-cnn-6-6'); const summary = await summarizer(documentText, { max_length: 130 }); ``` ### Option 2: **WebLLM** (For Advanced Models) ```javascript import * as webllm from "@mlc-ai/web-llm"; const engine = await webllm.CreateMLCEngine("Llama-2-7b-chat-hf-q4f32_1"); const reply = await engine.chat.completions.create({ messages: [{ role: "user", content: "Summarize this..." }] }); ``` ### Option 3: **MLC LLM** (Mobile Apps) ```javascript import * as MLCEngine from 'mlc-llm-react-native'; const engine = await MLCEngine.createEngine({ model: "Llama-2-7b-chat-q4f16_1", device: "auto" }); ``` ### Option 4: **RunAnywhere SDK** (Alternative) ```javascript import { RunAnywhere } from '@runanywhere/sdk'; const ai = new RunAnywhere({ modelType: 'llama-3-8b', offlineMode: true, privacy: 'on-device' }); ``` 📚 **See [LOCAL_AI_OPTIONS.md](./LOCAL_AI_OPTIONS.md) for complete comparison and implementation guide** ### 🎯 Try the Demo! Open `local-ai-demo.html` in your browser to see Transformers.js in action: - ✅ Works 100% offline (after initial model download) - ✅ Document summarization - ✅ Email drafting - ✅ Sentiment analysis - ✅ Code generation - ✅ Task extraction ## 📄 License MIT License - See LICENSE file for details ## 👥 Contributing Contributions welcome! Please read CONTRIBUTING.md first. ## 📧 Support For support, email [email protected] or open an issue. --- **Built with ❤️ for Enterprise Privacy** # Site: https://alnsight.onrender.com