ваше сообщение коммита
This commit is contained in:
1196
docs-en/FAQ.md
Normal file
1196
docs-en/FAQ.md
Normal file
File diff suppressed because it is too large
Load Diff
29
docs-en/README.md
Normal file
29
docs-en/README.md
Normal file
@@ -0,0 +1,29 @@
|
||||
# 📚 Digital Legal Entity Documentation
|
||||
|
||||
**🌐 Language / Язык:** [🇷🇺 Русский](../docs/README.md) | [🇬🇧 English](README.md)
|
||||
|
||||
This section contains materials about the DLE product: overview, business cases, technical specifications, and implementation instructions. Use the list below to quickly find the document you need.
|
||||
|
||||
## 📄 Available Documents
|
||||
|
||||
| File | Brief Description |
|
||||
| --- | --- |
|
||||
| [application-description.md](./application-description.md) | Overview of the platform's purpose, key benefits, and economic impact. |
|
||||
| [ai-assistant.md](./ai-assistant.md) | Detailed description of AI assistant capabilities and business cases. |
|
||||
| [setup-ai-assistant.md](./setup-ai-assistant.md) | Step-by-step setup of a local AI assistant with vector search. |
|
||||
| [setup-instruction.md](./setup-instruction.md) | Main installation and initial configuration instructions. |
|
||||
| [blockchain-for-business.md](./blockchain-for-business.md) | Business justification and use cases for blockchain in DLE. |
|
||||
| [blockchain-integration-technical.md](./blockchain-integration-technical.md) | Technical architecture of blockchain integration and smart contracts. |
|
||||
| [security.md](./security.md) | Security model, access control, and protection mechanisms. |
|
||||
| [service-terms.md](./service-terms.md) | Detailed terms for license purchase and maintenance. |
|
||||
| [FAQ.md](./FAQ.md) | Answers to frequently asked questions about licensing, installation, and features. |
|
||||
| [system-messages-management.md](./system-messages-management.md) | Technical specification for system message management. |
|
||||
| [tables-system.md](./tables-system.md) | Description of the electronic tables system, architecture, and integrations. |
|
||||
|
||||
---
|
||||
|
||||
### 🔗 Related Materials
|
||||
|
||||
- Legal documents and requirements are available in the [`../legal-en`](../legal-en/README.md) directory.
|
||||
- Updates and additional information are published in the root [`README`](../README.md).
|
||||
|
||||
387
docs-en/ai-assistant.md
Normal file
387
docs-en/ai-assistant.md
Normal file
@@ -0,0 +1,387 @@
|
||||
# DLE AI Assistant - Complete Business Capabilities
|
||||
|
||||
> **Philosophy**: AI doesn't replace people, but becomes their co-pilot, freeing up 85% of time for strategy and creativity.
|
||||
|
||||
## 📋 Table of Contents
|
||||
|
||||
1. [Introduction](#introduction)
|
||||
2. [Technology and Architecture](#technology-and-architecture)
|
||||
3. [Complete AI Assistant Capabilities](#complete-ai-assistant-capabilities)
|
||||
4. [Economic Impact](#economic-impact)
|
||||
5. [Setup and Integration](#setup-and-integration)
|
||||
|
||||
---
|
||||
|
||||
## Introduction
|
||||
|
||||
### What is AI Assistant in DLE?
|
||||
|
||||
**AI Assistant** is a built-in intelligent assistant that uses:
|
||||
- 🧠 **Ollama** - local open-source AI models (qwen2.5:7b)
|
||||
- 🔍 **Vector Search** - FAISS for semantic search
|
||||
- 📚 **RAG** (Retrieval-Augmented Generation) - knowledge base search
|
||||
- 💾 **Caching** - instant responses to frequent questions
|
||||
|
||||
### Key Benefits
|
||||
|
||||
1. **🏠 100% Local** - all data on your server
|
||||
2. **🔒 Complete Privacy** - nothing goes to the cloud
|
||||
3. **⚡ Lightning-fast Responses** - caching and vector search
|
||||
4. **💰 Zero API Costs** - no token payments
|
||||
5. **📈 Learns from Your Data** - personalized responses
|
||||
6. **🌍 Multilingual** - supports Russian, English, and more
|
||||
7. **🔄 24/7 Availability** - works around the clock
|
||||
|
||||
---
|
||||
|
||||
## Technology and Architecture
|
||||
|
||||
### How It Works?
|
||||
|
||||
```
|
||||
┌─────────────────────────────────────────────────────────┐
|
||||
│ AI Assistant Architecture │
|
||||
└─────────────────────────────────────────────────────────┘
|
||||
|
||||
┌─────────────────────────────────────────┐
|
||||
│ 1. User asks a question │
|
||||
│ "How to return a product?" │
|
||||
└────────────────┬────────────────────────┘
|
||||
↓
|
||||
┌─────────────────────────────────────────┐
|
||||
│ 2. Question Vectorization │
|
||||
│ Ollama mxbai-embed-large:latest │
|
||||
│ Text → Vector [768 numbers] │
|
||||
└────────────────┬────────────────────────┘
|
||||
↓
|
||||
┌─────────────────────────────────────────┐
|
||||
│ 3. Vector Search (FAISS) │
|
||||
│ Search for similar questions │
|
||||
│ Top-3 results by semantics │
|
||||
└────────────────┬────────────────────────┘
|
||||
↓
|
||||
┌─────────────────────────────────────────┐
|
||||
│ 4. Contextual Search (RAG) │
|
||||
│ Extract answers from knowledge base │
|
||||
│ Filter by tags/products │
|
||||
└────────────────┬────────────────────────┘
|
||||
↓
|
||||
┌─────────────────────────────────────────┐
|
||||
│ 5. Answer Generation (LLM) │
|
||||
│ Ollama qwen2.5:7b │
|
||||
│ Context + Question → Answer │
|
||||
└────────────────┬────────────────────────┘
|
||||
↓
|
||||
┌─────────────────────────────────────────┐
|
||||
│ 6. Caching │
|
||||
│ Save to cache for 1 hour │
|
||||
│ Next same question < 50ms │
|
||||
└────────────────┬────────────────────────┘
|
||||
↓
|
||||
┌─────────────────────────────────────────┐
|
||||
│ 7. Response to User │
|
||||
│ "Product can be returned within..." │
|
||||
└─────────────────────────────────────────┘
|
||||
```
|
||||
|
||||
### Technology Stack
|
||||
|
||||
| Component | Technology | Purpose |
|
||||
|-----------|------------|---------|
|
||||
| **LLM Model** | Ollama qwen2.5:7b | Text generation, dialogue |
|
||||
| **Embedding Model** | mxbai-embed-large | Text vectorization |
|
||||
| **Vector DB** | FAISS | Fast semantic search |
|
||||
| **Main DB** | PostgreSQL | Knowledge base storage |
|
||||
| **Cache** | Node.js Map + TTL | Accelerate repeated requests |
|
||||
| **Queue** | AI Queue | Priority task processing |
|
||||
|
||||
---
|
||||
|
||||
## Complete AI Assistant Capabilities
|
||||
|
||||
### 1. 📞 Customer Service and Support
|
||||
|
||||
**Automatic responses to common questions:**
|
||||
- Instant 24/7 responses via Telegram, Email, Web chat
|
||||
- Automatic problem diagnosis by order number
|
||||
- Ticket creation in CRM with prioritization
|
||||
- Knowledge base with vector search
|
||||
- Multilingual customer support from different countries
|
||||
- Real-time customer sentiment analysis
|
||||
- Escalation of complex questions to live operators
|
||||
- Personalized recommendations based on purchase history
|
||||
|
||||
**Messenger Integration:**
|
||||
- Telegram bot with AI assistant
|
||||
- Email support with automatic analysis
|
||||
- Web chat widget on website
|
||||
- SMS notifications and alerts
|
||||
- Web feedback forms
|
||||
|
||||
### 2. 📊 Analytics and Business Intelligence
|
||||
|
||||
**Automatic Data Analysis:**
|
||||
- Big data analysis and trend identification
|
||||
- Sales and demand forecasting for products/services
|
||||
- Customer segmentation by behavior and preferences
|
||||
- Marketing campaign effectiveness analysis
|
||||
- Financial indicator anomaly detection
|
||||
- Interactive dashboards with key metrics
|
||||
- Competitive analysis
|
||||
- Benchmarking and industry comparisons
|
||||
|
||||
### 3. 💰 Financial Planning and Accounting
|
||||
|
||||
**Financial Process Automation:**
|
||||
- Automatic accounting and report generation
|
||||
- Cash flow forecasting and budget planning
|
||||
- Product and service profitability analysis
|
||||
- Financial risk identification and mitigation recommendations
|
||||
- Automatic invoice creation and payment tracking
|
||||
- Tax planning optimization
|
||||
- KPI and financial indicator monitoring
|
||||
- Consolidated reporting across all departments
|
||||
|
||||
### 4. 🎯 Marketing and Sales
|
||||
|
||||
**Content Marketing:**
|
||||
- Content generation for social media, blogs, email campaigns
|
||||
- Marketing message personalization for each customer
|
||||
- A/B testing of various advertising options
|
||||
- Customer acquisition channel effectiveness analysis
|
||||
- Automatic email campaign creation
|
||||
- SEO content and meta tag optimization
|
||||
- Online reputation management
|
||||
|
||||
**Sales:**
|
||||
- Automatic commercial proposal creation
|
||||
- Customer base analysis and opportunity identification
|
||||
- Personalized product recommendations
|
||||
- Automatic sales funnel tracking
|
||||
- Cold call script generation
|
||||
|
||||
### 5. 👥 HR and Personnel Management
|
||||
|
||||
**Recruitment and Personnel Management:**
|
||||
- Automatic resume screening and candidate pre-selection
|
||||
- Interviewing with chatbot
|
||||
- Employee performance analysis
|
||||
- Training and development planning
|
||||
- Automatic job description creation
|
||||
- Employee satisfaction analysis
|
||||
- Vacation and shift planning
|
||||
- Team mood monitoring
|
||||
|
||||
**Employee Training:**
|
||||
- Personalized training programs for employees
|
||||
- Interactive course and material creation
|
||||
- Adaptive learning for individual needs
|
||||
- Knowledge and skill assessment
|
||||
- Career development recommendations
|
||||
- Knowledge base and FAQ creation
|
||||
|
||||
### 6. 📋 Project and Task Management
|
||||
|
||||
**Project Work Automation:**
|
||||
- Automatic project planning and resource allocation
|
||||
- Progress tracking and delay identification
|
||||
- Workflow optimization and bottleneck elimination
|
||||
- Automatic project reporting
|
||||
- Risk management and reserve planning
|
||||
- Distributed team coordination
|
||||
- Management methodology effectiveness analysis
|
||||
|
||||
### 7. 🔍 Research and Development
|
||||
|
||||
**Analytics and Research:**
|
||||
- Market and competitive environment analysis
|
||||
- New technology and trend research
|
||||
- New product/service idea generation
|
||||
- Patent and intellectual property analysis
|
||||
- Technical documentation and specifications
|
||||
- Customer feedback analysis for product improvement
|
||||
|
||||
### 8. 🛡️ Security and Compliance
|
||||
|
||||
**Security Monitoring:**
|
||||
- IT system security monitoring
|
||||
- Suspicious activity and threat detection
|
||||
- Automatic security policy updates
|
||||
- Regulatory compliance analysis
|
||||
- Access and user rights management
|
||||
- Log auditing and incident investigation
|
||||
|
||||
### 9. 📈 Operational Efficiency
|
||||
|
||||
**Process Optimization:**
|
||||
- Supply chain and logistics optimization
|
||||
- Production planning and inventory management
|
||||
- Routine process automation
|
||||
- Product and service quality analysis
|
||||
- Asset management and maintenance
|
||||
- Energy and resource consumption optimization
|
||||
|
||||
### 10. 🎓 Training and Development
|
||||
|
||||
**Corporate Training:**
|
||||
- Personalized training programs for employees
|
||||
- Interactive course and material creation
|
||||
- Adaptive learning for individual needs
|
||||
- Knowledge and skill assessment
|
||||
- Career development recommendations
|
||||
- Knowledge base and FAQ creation
|
||||
|
||||
### 11. 🌐 International Activities
|
||||
|
||||
**Global Capabilities:**
|
||||
- Automatic document and communication translation
|
||||
- Target market cultural analysis
|
||||
- Content adaptation for local requirements
|
||||
- International legislation change monitoring
|
||||
- Currency risk management
|
||||
- Coordination with foreign partners
|
||||
|
||||
### 12. 🔮 Strategic Planning
|
||||
|
||||
**Long-term Planning:**
|
||||
- Macroeconomic trend analysis and business impact
|
||||
- Scenario planning for various situations
|
||||
- Merger and acquisition opportunity analysis
|
||||
- Investment project evaluation
|
||||
- New market entry planning
|
||||
- Long-term development strategy development
|
||||
|
||||
### 13. 📱 Mobile and Digital Solutions
|
||||
|
||||
**Digital Transformation:**
|
||||
- Mobile app development with AI features
|
||||
- Chatbot creation for various platforms
|
||||
- Voice assistants for business processes
|
||||
- IoT device integration
|
||||
- Social media automation
|
||||
- Virtual assistant creation
|
||||
|
||||
### 14. 🏭 Industry Solutions
|
||||
|
||||
**Specialized Solutions:**
|
||||
- **Healthcare**: diagnosis, treatment planning, patient management
|
||||
- **Finance**: credit scoring, risk management, trading algorithms
|
||||
- **Retail**: personalization, inventory management, pricing
|
||||
- **Manufacturing**: quality control, predictive maintenance, optimization
|
||||
- **Real Estate**: valuation, market analysis, property management
|
||||
- **Education**: adaptive learning, knowledge assessment, planning
|
||||
|
||||
### 15. 🔄 Integration and Automation
|
||||
|
||||
**System Integration:**
|
||||
- API integrations with external systems
|
||||
- Workflow automation (RPA)
|
||||
- Data synchronization between various platforms
|
||||
- Unified ecosystem creation for all business processes
|
||||
- Data and system migration
|
||||
- Integrated solution performance monitoring
|
||||
|
||||
---
|
||||
|
||||
## Economic Impact
|
||||
|
||||
### Comparison with Cloud Solutions
|
||||
|
||||
| Characteristic | DLE AI (Local) | ChatGPT API | Claude API |
|
||||
|----------------|----------------|-------------|------------|
|
||||
| **Cost** | $0 | ~$0.02/request | ~$0.03/request |
|
||||
| **Privacy** | ✅ 100% | ❌ Data in OpenAI | ❌ Data in Anthropic |
|
||||
| **Speed (cache)** | 50ms | 500-2000ms | 500-2000ms |
|
||||
| **Offline Work** | ✅ Yes | ❌ No | ❌ No |
|
||||
| **Business Customization** | ✅ Full | ⚠️ Limited | ⚠️ Limited |
|
||||
| **API Limits** | ❌ None | ✅ Yes | ✅ Yes |
|
||||
|
||||
### Savings with Cloud API Usage
|
||||
|
||||
**At 10,000 requests/month:**
|
||||
- ChatGPT API: $200/month = **$2,400/year**
|
||||
- Claude API: $300/month = **$3,600/year**
|
||||
- DLE AI: **$0/year** ✅
|
||||
|
||||
**At 50,000 requests/month:**
|
||||
- ChatGPT API: $1,000/month = **$12,000/year**
|
||||
- Claude API: $1,500/month = **$18,000/year**
|
||||
- DLE AI: **$0/year** ✅
|
||||
|
||||
**At 100,000 requests/month:**
|
||||
- ChatGPT API: $2,000/month = **$24,000/year**
|
||||
- Claude API: $3,000/month = **$36,000/year**
|
||||
- DLE AI: **$0/year** ✅
|
||||
|
||||
### Total Savings Across All Areas
|
||||
|
||||
| Area | Annual Savings | ROI |
|
||||
|------|----------------|-----|
|
||||
| **Customer Service** | $57,600 | 5,760% |
|
||||
| **Supplier Work + AI Purchaser** | $64,800 | 6,480% |
|
||||
| **Partner Work** | $43,200 | 4,320% |
|
||||
| **Employee Training** | $30,000 | 3,000% |
|
||||
| **Personnel Management** | $57,600 | 5,760% |
|
||||
| **Reporting and Analysis** | $144,000 | 14,400% |
|
||||
| **Content Marketing** | $86,400 | 8,640% |
|
||||
| **API Token Savings** | $24,000-36,000 | 2,400-3,600% |
|
||||
| **TOTAL:** | **$507,600-519,600** | **50,760-51,960%** |
|
||||
|
||||
**DLE Implementation Cost**: $1,000 (one-time, Standard Support)
|
||||
**ROI**: 50,760-51,960% or **507-519x return on investment**
|
||||
|
||||
### Additional Benefits
|
||||
|
||||
**Immeasurable Advantages:**
|
||||
- 📈 Improved work quality
|
||||
- 😊 Customer satisfaction +19%
|
||||
- 👥 Employee retention +15%
|
||||
- 🚀 Decision speed +300%
|
||||
- 💡 More time for strategy
|
||||
- 🎯 Less routine, more creativity
|
||||
|
||||
---
|
||||
|
||||
## Setup and Integration
|
||||
|
||||
**What You Need to Launch:**
|
||||
- ✅ Install DLE (see [Installation Instructions](./setup-instruction.md))
|
||||
- ✅ Run Ollama and install models
|
||||
- ✅ Create knowledge base (FAQ, suppliers, HR)
|
||||
- ✅ Configure AI assistant for your business
|
||||
- ✅ Integrate with Telegram/Email (optional)
|
||||
|
||||
**Setup Time:**
|
||||
- 🚀 **Quick Start**: 20-30 minutes (basic FAQ)
|
||||
- ⚙️ **Full Setup**: 1-2 hours (all capabilities)
|
||||
|
||||
**Result:**
|
||||
- 🤖 Working AI assistant 24/7
|
||||
- 📚 Knowledge base for customer responses
|
||||
- 🔍 Vector search on your data
|
||||
- 💰 Savings up to $519,600/year
|
||||
|
||||
### Next Steps
|
||||
|
||||
1. 📖 [Install DLE](./setup-instruction.md)
|
||||
2. 🤖 [Configure AI Assistant](./setup-ai-assistant.md)
|
||||
3. 📚 Upload knowledge base
|
||||
4. 🚀 Start saving time and money!
|
||||
|
||||
---
|
||||
|
||||
## Support
|
||||
|
||||
**Need Help with AI Setup?**
|
||||
|
||||
📧 **Email**: info@hb3-accelerator.com
|
||||
💬 **Chat**: https://hb3-accelerator.com
|
||||
📚 **Documentation**: [FAQ](./FAQ.md)
|
||||
🎓 **Training**: Online sessions for token holders
|
||||
|
||||
---
|
||||
|
||||
**© 2024-2025 Tarabanov Alexander Viktorovich. All rights reserved.**
|
||||
|
||||
**Last Updated**: October 2025
|
||||
|
||||
333
docs-en/application-description.md
Normal file
333
docs-en/application-description.md
Normal file
@@ -0,0 +1,333 @@
|
||||
# Digital Legal Entity (DLE) - Web Application for Business Management
|
||||
|
||||
> **Full Control. Zero Subscription Fee. AI Saves $507,600-519,600/year.**
|
||||
|
||||
---
|
||||
|
||||
## 📋 Table of Contents
|
||||
|
||||
1. [Who This Application Is For](#who-this-application-is-for)
|
||||
2. [Key Benefits](#key-benefits)
|
||||
3. [Main Features](#main-features)
|
||||
4. [Economic Impact](#economic-impact)
|
||||
5. [Technical Details](#technical-details)
|
||||
6. [Purchase Terms](#purchase-terms)
|
||||
7. [Additional Documentation](#additional-documentation)
|
||||
|
||||
---
|
||||
|
||||
## 🎯 Who This Application Is For
|
||||
|
||||
DLE is created for **all types of organizations** that need transparent collective management:
|
||||
|
||||
- **Commercial organizations** (LLC, JSC, sole proprietors, holdings)
|
||||
- **Non-profit organizations** (NPOs, foundations, associations)
|
||||
- **Government structures** (municipalities, agencies)
|
||||
- **Investment funds** and **venture companies**
|
||||
- **Startups** and **small businesses**
|
||||
- **Cooperatives** and **associations**
|
||||
|
||||
---
|
||||
|
||||
## 💎 Key Benefits
|
||||
|
||||
### Why DLE, Not SaaS?
|
||||
|
||||
| Parameter | DLE | Regular SaaS |
|
||||
|----------|-----|--------------|
|
||||
| **💰 Cost** | $1,000 one-time | $200-500/month |
|
||||
| **📊 Data Ownership** | 100% yours | With provider |
|
||||
| **🔧 Customization** | Full | Limited |
|
||||
| **🌍 Legal Compliance** | For your region | Depends on provider |
|
||||
| **🛡️ Regulatory Compliance** | GDPR, CCPA, 152-FZ | Limited |
|
||||
| **🤖 AI Without Limits** | Free | $200+/month |
|
||||
| **🔗 Blockchain** | Built-in | None or extra fee |
|
||||
| **⚡ Updates** | Free 5 years (for token holders) | Depends on plan |
|
||||
|
||||
### 🎁 What You Get
|
||||
|
||||
✅ **Perpetual License** - pay once, use forever
|
||||
✅ **Free Updates** - all improvements without extra payment
|
||||
✅ **Full Source Code** - complete transparency
|
||||
✅ **AI on Your Server** - 100% confidentiality
|
||||
✅ **Blockchain Protection** - assets cannot be stolen
|
||||
✅ **Asset Tokenization** - 99% savings on transactions
|
||||
✅ **Voting Rights** - influence on product development
|
||||
✅ **Ready Regulatory Package** - privacy policy, personal data consent, user agreement, cookie policy
|
||||
✅ **70% Refund Guarantee** - within 5 years if no updates
|
||||
|
||||
---
|
||||
|
||||
## 🚀 Main Features
|
||||
|
||||
### 1. CRM and Contact Management
|
||||
|
||||
**Full Customer Management System**:
|
||||
- Centralized contact database with interaction history
|
||||
- Grouping by tags and categories
|
||||
- Task and reminder system
|
||||
- Import from CSV, Excel, external CRMs
|
||||
- Export in various formats
|
||||
- Document and template builder
|
||||
|
||||
### 2. AI Assistant with Vector Search
|
||||
|
||||
**Local AI on Your Server**:
|
||||
- Ollama + Vector Search (RAG)
|
||||
- qwen2.5:7b model (or any other)
|
||||
- 100% confidentiality (data doesn't leave)
|
||||
- Zero API token costs
|
||||
|
||||
**Key Capabilities**:
|
||||
- **📞 Customer Service** - 24/7 responses, customer sentiment analysis
|
||||
- **📊 Analytics and Business Intelligence** - data analysis, sales forecasting
|
||||
- **💰 Financial Planning** - automated accounting, profitability analysis
|
||||
- **🎯 Marketing and Sales** - content generation, message personalization
|
||||
- **👥 HR and Personnel Management** - resume screening, training planning
|
||||
- **📋 Project Management** - automatic planning, progress tracking
|
||||
- **🔍 Research and Development** - market analysis, idea generation
|
||||
- **🛡️ Security and Compliance** - security monitoring, compliance analysis
|
||||
- **📈 Operational Efficiency** - process optimization, automation
|
||||
- **🎓 Training and Development** - personalized training programs
|
||||
- **🌐 International Activity** - automatic translation, content adaptation
|
||||
- **🔮 Strategic Planning** - trend analysis, scenario planning
|
||||
- **📱 Mobile Solutions** - app development, chatbot creation
|
||||
- **🏭 Industry Solutions** - specialized solutions for different industries
|
||||
- **🔄 Integration and Automation** - API integrations, process automation
|
||||
|
||||
**Economic Impact**: $507,600-519,600/year with ROI 50,760-51,960%
|
||||
|
||||
> 💡 **Detailed Information**: See [AI Assistant - Full Description](./ai-assistant.md) - detailed description of all capabilities, use cases, architecture, and savings calculations.
|
||||
|
||||
### 3. Omnichannel Communications
|
||||
|
||||
**Unified Interface for All Channels**:
|
||||
- 💬 Telegram bot, 📧 Email, 🌐 Web chat, 📱 SMS, 📋 Web forms
|
||||
- One context for customer across all channels
|
||||
- AI auto-responses trained on your data
|
||||
- Analytics by channel and response time
|
||||
|
||||
### 4. Blockchain Management and Tokenization
|
||||
|
||||
**Smart Contracts for Business**:
|
||||
- Tokenization of any assets (real estate, IP, shares)
|
||||
- Management through token holder voting
|
||||
- Multi-chain support (7+ networks: Ethereum, Polygon, BSC, Arbitrum, Optimism, Avalanche, Base)
|
||||
- Transparency of all operations on blockchain
|
||||
|
||||
> 💡 **Detailed Information**: See [Blockchain for Business](./blockchain-for-business.md) - practical use cases and [Blockchain Technical Documentation](./blockchain-integration-technical.md) - for developers.
|
||||
|
||||
### 5. Groups and Team Spaces
|
||||
|
||||
- Customizable spaces for projects
|
||||
- Detailed permission system (20+ types)
|
||||
- Roles: Editor, ReadOnly, User
|
||||
- Integration with CRM and communications
|
||||
|
||||
### 6. Internal Tools
|
||||
|
||||
- Electronic tables (Excel-like)
|
||||
- Analytics and reporting system
|
||||
- Metrics monitoring
|
||||
- WebSSH for server management
|
||||
|
||||
### 7. Security and Protection
|
||||
|
||||
**Multi-Level DLE Protection**:
|
||||
- **Network Security** - TLS 1.3, HTTPS, DDoS protection
|
||||
- **Authentication** - SIWE (wallet login), sessions in DB
|
||||
- **Authorization** - detailed permissions (20+ types), token gating
|
||||
- **Data Protection** - AES-256 encryption, SQL injection protection
|
||||
- **Blockchain Protection** - tokens cannot be stolen without voting
|
||||
- **Monitoring** - logging, alerts, audit trail
|
||||
|
||||
**Regulatory Compliance**:
|
||||
- **GDPR** - compliance with European personal data protection requirements
|
||||
- **CCPA** - compliance with California requirements
|
||||
- **Russian Legislation** - compliance with 152-FZ "On Personal Data"
|
||||
- **Local Storage** - all data remains on your server
|
||||
- **Transparency** - full control over personal data processing
|
||||
|
||||
> 💡 **Detailed Information**: See [DLE Security](./security.md) - detailed description of all protection levels, attack scenarios, security recommendations, and regulatory compliance.
|
||||
|
||||
## 💰 Economic Impact
|
||||
|
||||
### Comparison with SaaS Over 5 Years
|
||||
|
||||
**Typical SaaS Stack**:
|
||||
```
|
||||
CRM (HubSpot): $200/mo × 60 mo = $12,000
|
||||
Chatbot (Intercom): $150/mo × 60 mo = $9,000
|
||||
Email (SendGrid): $100/mo × 60 mo = $6,000
|
||||
AI (ChatGPT API): $200/mo × 60 mo = $12,000
|
||||
─────────────────────────────
|
||||
TOTAL: $39,000
|
||||
```
|
||||
|
||||
**DLE**:
|
||||
```
|
||||
License (one-time): $1,000
|
||||
Updates (5 years): $0
|
||||
AI without limits: $0
|
||||
─────────────────────────────
|
||||
TOTAL: $1,000
|
||||
|
||||
SAVINGS: $38,000
|
||||
```
|
||||
|
||||
### ROI from AI Assistant
|
||||
|
||||
**Comparison with Cloud Solutions**:
|
||||
|
||||
| Characteristic | DLE AI (Local) | ChatGPT API | Claude API |
|
||||
|----------------|-------------------|-------------|------------|
|
||||
| **Cost** | $0 | ~$0.02/request | ~$0.03/request |
|
||||
| **Confidentiality** | ✅ 100% | ❌ Data in OpenAI | ❌ Data in Anthropic |
|
||||
| **Speed (cache)** | 50ms | 500-2000ms | 500-2000ms |
|
||||
| **Offline Work** | ✅ Yes | ❌ No | ❌ No |
|
||||
| **Business Customization** | ✅ Full | ⚠️ Limited | ⚠️ Limited |
|
||||
| **API Limits** | ❌ None | ✅ Yes | ✅ Yes |
|
||||
|
||||
**Total Savings Across All Areas**:
|
||||
- Customer Service: $57,600
|
||||
- Supplier Relations + AI Purchaser: $64,800
|
||||
- Partner Relations: $43,200
|
||||
- Staff Training: $30,000
|
||||
- Personnel Management: $57,600
|
||||
- Reporting and Analysis: $144,000
|
||||
- Content Marketing: $86,400
|
||||
- API Token Savings: $24,000-36,000
|
||||
- **TOTAL**: $507,600-519,600/year
|
||||
|
||||
**Investment**: $1,000
|
||||
**ROI**: 50,760-51,960% or 507-519x return ✅
|
||||
|
||||
### Savings on Asset Tokenization
|
||||
|
||||
**Traditional Real Estate Transaction ($1M)**:
|
||||
- Realtor Commission: 3-5% = $30,000-50,000
|
||||
- Legal Services: $5,000-10,000
|
||||
- Title Registration: $2,000-5,000
|
||||
- Time: 2-6 months
|
||||
- **Total**: $37,000-65,000
|
||||
|
||||
**Tokenized Transaction via DLE**:
|
||||
- Gas Fees: $50-200
|
||||
- Smart Contract (one-time): $500
|
||||
- Time: 5-10 minutes
|
||||
- **Total**: $550-700
|
||||
|
||||
**Savings**: 99% costs and 99% time ✅
|
||||
|
||||
---
|
||||
|
||||
## 🔧 Technical Details
|
||||
|
||||
### Architecture
|
||||
|
||||
**Modern Technology Stack**:
|
||||
- **Frontend**: Vue.js 3 + Vite + Element Plus
|
||||
- **Backend**: Node.js + Express
|
||||
- **Database**: PostgreSQL + pgvector
|
||||
- **AI**: Ollama (qwen2.5:7b) + FAISS Vector Search
|
||||
- **Blockchain**: Ethers.js v6 + Hardhat
|
||||
- **Containerization**: Docker Compose
|
||||
|
||||
### Design Principles
|
||||
|
||||
✅ **Microservices Architecture** - easy to scale
|
||||
✅ **Full Control** - you own the application and data
|
||||
✅ **Modularity** - add new features through modules
|
||||
✅ **Security** - multi-level protection (TLS, AES-256, CSRF)
|
||||
✅ **API-first** - REST + WebSocket for integrations
|
||||
|
||||
### Deployment
|
||||
|
||||
**Flexible Installation Options**:
|
||||
- **On Your Server** - full control
|
||||
- **In Cloud** - AWS, Google Cloud, Azure
|
||||
- **On Local Infrastructure** - regulatory compliance
|
||||
- **Docker Compose** - deployment with one command
|
||||
|
||||
**Minimum Requirements**:
|
||||
- **CPU**: 4 cores
|
||||
- **RAM**: 12 GB (4 GB application + 6 GB AI + 2 GB Vector Search)
|
||||
- **Storage**: 100 GB SSD
|
||||
- **OS**: Ubuntu 20.04+, Debian 11+, CentOS 8+, any Linux with Docker
|
||||
|
||||
> 💡 **Detailed Information**: See [Installation Instructions](./setup-instruction.md) - step-by-step setup and [AI Assistant Setup](./setup-ai-assistant.md) - specialized installation.
|
||||
|
||||
---
|
||||
|
||||
## 💳 Purchase Terms
|
||||
|
||||
> 💡 **Detailed Information**: See [Service Terms](./service-terms.md) - detailed description of licensing model, pricing, guarantees, and purchase process.
|
||||
|
||||
### Quick Overview
|
||||
|
||||
**Perpetual License**:
|
||||
- 💰 **Pay Once** - use forever
|
||||
- 📄 **Full Source Code** with documentation
|
||||
- 🔄 **Free Updates** 5 years for license token holders (see [legal-en/service-terms.md](../legal-en/service-terms.md))
|
||||
- 🗳️ **Voting Rights** for product development
|
||||
|
||||
**Pricing**:
|
||||
| Package | Price | Votes | Support |
|
||||
|-------|------|--------|-----------|
|
||||
| **Standard Support** | $1,000 USDT | 1 vote | 24/7 chat |
|
||||
| **Priority Support** | $10,000 USDT | 10 votes | 24/7 chat |
|
||||
|
||||
**What's Included**:
|
||||
✅ Perpetual license for use
|
||||
✅ Source code with full documentation
|
||||
✅ Free updates 5 years for token holders
|
||||
✅ Technical support
|
||||
✅ Governance tokens on blockchain
|
||||
✅ Ready regulatory package
|
||||
✅ 70% refund guarantee within 5 years
|
||||
|
||||
**How to Purchase**:
|
||||
1. Contact us: info@hb3-accelerator.com
|
||||
2. Receive payment details (USDT)
|
||||
3. Pay for license
|
||||
4. Receive code access and DLE tokens
|
||||
|
||||
---
|
||||
|
||||
## 📚 Additional Documentation
|
||||
|
||||
### 🎯 Quick Start
|
||||
- 📋 **[FAQ](./FAQ.md)** - answers to 50+ questions (11 sections)
|
||||
- 🔧 **[Installation](./setup-instruction.md)** - step-by-step setup in 15 minutes
|
||||
- 🤖 **[AI Setup](./setup-ai-assistant.md)** - launch AI assistant
|
||||
|
||||
### 💼 Business and Economics
|
||||
- 💰 **[Blockchain for Business](./blockchain-for-business.md)** - asset tokenization, 99% savings on transactions
|
||||
- 🤖 **[AI Assistant](./ai-assistant.md)** - $507,600-519,600/year savings, "AI+Human" philosophy
|
||||
- 📊 **[Tables System](./tables-system.md)** - data and document management
|
||||
|
||||
### 🔧 Technical Documentation
|
||||
- 🔗 **[Blockchain Integration](./blockchain-integration-technical.md)** - smart contracts, API, code examples
|
||||
- 🛡️ **[Security](./security.md)** - multi-level protection, GDPR/CCPA/152-FZ compliance
|
||||
- ⚖️ **[Service Terms](./service-terms.md)** - licensing, support, guarantees
|
||||
|
||||
### 📖 General Information
|
||||
- 📖 **[Main README](../README.md)** - complete list of all documents
|
||||
- ⚖️ **[Legal Documentation](../legal-en/README.md)** - legal information
|
||||
|
||||
### 📞 Support and Contacts
|
||||
|
||||
**Main Contacts**:
|
||||
- 📧 **Email**: info@hb3-accelerator.com
|
||||
- 🌐 **Website**: https://hb3-accelerator.com
|
||||
- 💬 **Support Chat**: https://hb3-accelerator.com/
|
||||
- 🐙 **GitHub**: https://github.com/VC-HB3-Accelerator
|
||||
---
|
||||
|
||||
**© 2024-2025 Tarabanov Alexander Viktorovich. All rights reserved.**
|
||||
|
||||
**Digital Legal Entity (DLE)** - revolution in business management through blockchain and AI.
|
||||
|
||||
**Document Version**: 2.0.0
|
||||
**Last Update**: October 2025
|
||||
|
||||
1288
docs-en/blockchain-for-business.md
Normal file
1288
docs-en/blockchain-for-business.md
Normal file
File diff suppressed because it is too large
Load Diff
1049
docs-en/blockchain-integration-technical.md
Normal file
1049
docs-en/blockchain-integration-technical.md
Normal file
File diff suppressed because it is too large
Load Diff
1309
docs-en/security.md
Normal file
1309
docs-en/security.md
Normal file
File diff suppressed because it is too large
Load Diff
43
docs-en/service-terms.md
Normal file
43
docs-en/service-terms.md
Normal file
@@ -0,0 +1,43 @@
|
||||
<!--
|
||||
Copyright (c) 2024-2025 Tarabanov Alexander Viktorovich
|
||||
All rights reserved.
|
||||
|
||||
This software is proprietary and confidential.
|
||||
Unauthorized copying, modification, or distribution is prohibited.
|
||||
|
||||
For licensing inquiries: info@hb3-accelerator.com
|
||||
Website: https://hb3-accelerator.com
|
||||
GitHub: https://github.com/VC-HB3-Accelerator
|
||||
-->
|
||||
|
||||
# Digital Legal Entity Purchase and Service Terms — Brief Summary
|
||||
|
||||
> Important: This is a brief summary of terms. The canonical and legally binding full text is located here: [legal-en/service-terms.md](../legal-en/service-terms.md).
|
||||
|
||||
Key Points:
|
||||
- License Type: Perpetual, rights determined by number of tokens (1 or 10).
|
||||
- Updates and Basic Maintenance: Free for 5 years from on-chain token transfer date.
|
||||
- Development Voting: 1 token = 1 vote, decisions by majority ≥51%.
|
||||
- 70% Refund possible if program conditions are met (see original).
|
||||
- Support, releases, knowledge base — through application `https://hb3-accelerator.com/`.
|
||||
- Full list of limitations/liability — in original.
|
||||
|
||||
## 📚 Additional Documentation
|
||||
|
||||
### Explore DLE Capabilities
|
||||
- 🤖 **[AI Assistant](./ai-assistant.md)** - second pilot for team, frees time for strategy
|
||||
- 💼 **[Blockchain for Business](./blockchain-for-business.md)** - asset tokenization and business solutions
|
||||
- 🛡️ **[DLE Security](./security.md)** - multi-level protection for your business
|
||||
|
||||
### Technical Documentation
|
||||
- 🔗 **[Blockchain Technical Documentation](./blockchain-integration-technical.md)** - for developers
|
||||
- 🔧 **[Installation Instructions](./setup-instruction.md)** - step-by-step setup
|
||||
- 📝 **[Application Description](./application-description.md)** - functionality overview
|
||||
- 📋 **[FAQ](./FAQ.md)** - frequently asked questions
|
||||
|
||||
### Contacts and Support
|
||||
- 📧 **Email**: info@hb3-accelerator.com
|
||||
- 🌐 **Website**: https://hb3-accelerator.com
|
||||
- 💬 **Support Chat**: https://hb3-accelerator.com/
|
||||
- 🐙 **GitHub**: https://github.com/VC-HB3-Accelerator
|
||||
|
||||
929
docs-en/setup-ai-assistant.md
Normal file
929
docs-en/setup-ai-assistant.md
Normal file
@@ -0,0 +1,929 @@
|
||||
# AI Assistant Setup Guide with Vector Search
|
||||
|
||||
## 🤖 Complete Guide to Launching an Intelligent Assistant
|
||||
|
||||
This document describes the step-by-step process of setting up an AI assistant for solving business tasks through spreadsheets and vector search.
|
||||
|
||||
---
|
||||
|
||||
## 📚 What You'll Set Up
|
||||
|
||||
After completing the instructions, you will have:
|
||||
|
||||
✅ Working AI assistant with local model (Ollama)
|
||||
✅ Knowledge base for customer responses (FAQ)
|
||||
✅ Supplier work automation
|
||||
✅ Employee training system
|
||||
✅ Vector search on your data
|
||||
✅ Significant time and resource savings
|
||||
|
||||
> 💡 **Economic Impact**: See [AI Assistant](./ai-assistant.md) - detailed description of all capabilities and savings calculations.
|
||||
|
||||
---
|
||||
|
||||
## ⏱️ Setup Time
|
||||
|
||||
- **Quick Setup**: 20-30 minutes (basic FAQ)
|
||||
- **Full Setup**: 1-2 hours (all capabilities)
|
||||
|
||||
---
|
||||
|
||||
## Step 1: Installing and Running Ollama
|
||||
|
||||
### 1.1 Checking Ollama Status
|
||||
|
||||
1. Go to **Settings** → **Integrations** tab
|
||||
2. Find the **"Ollama"** block and click **"Details"**
|
||||
3. Check connection status:
|
||||
- ✅ **"Ollama is running"** — ready, proceed to step 1.3
|
||||
- ❌ **"Ollama API not responding"** — proceed to step 1.2
|
||||
|
||||
### 1.2 Starting Ollama (if not running)
|
||||
|
||||
If Ollama is not running, execute in terminal:
|
||||
|
||||
```bash
|
||||
# For Docker (recommended)
|
||||
docker-compose up -d ollama
|
||||
|
||||
# Or locally
|
||||
ollama serve
|
||||
```
|
||||
|
||||
Refresh the page and check status again.
|
||||
|
||||
### 1.3 Installing AI Model
|
||||
|
||||
1. In the **Ollama** section, click **"Install Model"**
|
||||
2. Select model:
|
||||
- **qwen2.5:7b** (recommended) — for Russian, 4.7 GB
|
||||
- **llama2:7b** — for English, 3.8 GB
|
||||
- **mistral:7b** — universal, 4.1 GB
|
||||
3. Click **"Install"**
|
||||
4. Wait for download to complete (5-10 minutes depending on internet speed)
|
||||
|
||||
> 💡 **Tip**: Model downloads once and is stored locally
|
||||
|
||||
### 1.4 Installing Embedding Model
|
||||
|
||||
1. In the same section, find **"Install Embedding Model"**
|
||||
2. Select model:
|
||||
- **mxbai-embed-large:latest** (recommended) — 670 MB
|
||||
- **nomic-embed-text:latest** — alternative, 274 MB
|
||||
3. Click **"Install"**
|
||||
|
||||
> ⚠️ **Important**: Embedding model is needed for vector search (RAG)
|
||||
|
||||
---
|
||||
|
||||
## Step 2: Creating Knowledge Base (Spreadsheets)
|
||||
|
||||
### 2.1 Creating FAQ Table
|
||||
|
||||
1. Go to **Tables** (in main menu)
|
||||
2. Click **"+ Create Table"**
|
||||
3. Fill in:
|
||||
- **Name**: `FAQ - Frequently Asked Questions`
|
||||
- **Description**: `Knowledge base for AI assistant customer service`
|
||||
4. Click **"Create"**
|
||||
|
||||
### 2.2 Configuring Table Columns
|
||||
|
||||
Add the following columns:
|
||||
|
||||
#### Column 1: Question (required for RAG)
|
||||
|
||||
1. Click **"+ Add Column"**
|
||||
2. Fill in:
|
||||
- **Name**: `Question`
|
||||
- **Type**: `Text`
|
||||
- **Purpose**: Select `Question for AI` ⭐
|
||||
3. Click **"Save"**
|
||||
|
||||
> ⚠️ **Critical**: Must select purpose "Question for AI" — this field will be indexed for vector search
|
||||
|
||||
#### Column 2: Answer (required for RAG)
|
||||
|
||||
1. Click **"+ Add Column"**
|
||||
2. Fill in:
|
||||
- **Name**: `Answer`
|
||||
- **Type**: `Text`
|
||||
- **Purpose**: Select `AI Answer` ⭐
|
||||
3. Click **"Save"**
|
||||
|
||||
#### Column 3: Product (optional, for filtering)
|
||||
|
||||
1. Click **"+ Add Column"**
|
||||
2. Fill in:
|
||||
- **Name**: `Product`
|
||||
- **Type**: `Multiple Choice`
|
||||
- **Options**: `Basic`, `Premium`, `Corporate`
|
||||
- **Purpose**: Select `Product Filter`
|
||||
3. Click **"Save"**
|
||||
|
||||
#### Column 4: Tags (optional, for categorization)
|
||||
|
||||
1. Click **"+ Add Column"**
|
||||
2. Fill in:
|
||||
- **Name**: `Tags`
|
||||
- **Type**: `Multiple Choice`
|
||||
- **Options**: `Payment`, `Delivery`, `Return`, `Warranty`, `Support`
|
||||
- **Purpose**: Select `User Tags`
|
||||
3. Click **"Save"**
|
||||
|
||||
#### Column 5: Priority (optional)
|
||||
|
||||
1. Click **"+ Add Column"**
|
||||
2. Fill in:
|
||||
- **Name**: `Priority`
|
||||
- **Type**: `Number`
|
||||
- **Purpose**: Select `Priority`
|
||||
3. Click **"Save"**
|
||||
|
||||
> 💡 **Tip**: Questions with higher priority will be shown to AI first
|
||||
|
||||
### 2.3 Filling Knowledge Base
|
||||
|
||||
Add typical questions and answers:
|
||||
|
||||
**Example 1: Payment**
|
||||
|
||||
| Question | Answer | Product | Tags | Priority |
|
||||
|----------|--------|---------|------|----------|
|
||||
| How to pay for order? | We accept payment by bank card, PayPal, or bank transfer. Choose convenient method when placing order. | All | Payment | 10 |
|
||||
| Can I pay in installments? | Yes, for orders from 50,000₽ installment plan available for 3, 6 or 12 months without overpayment. | Premium, Corporate | Payment | 8 |
|
||||
|
||||
**Example 2: Delivery**
|
||||
|
||||
| Question | Answer | Product | Tags | Priority |
|
||||
|----------|--------|---------|------|----------|
|
||||
| How long does delivery take? | Standard delivery: 3-5 business days in Russia. Express delivery: 1-2 days in major cities. | All | Delivery | 10 |
|
||||
| How much does delivery cost? | Free delivery for orders from 5,000₽. For orders less than 5,000₽ delivery cost is 300₽. | All | Delivery | 9 |
|
||||
|
||||
**Example 3: Return**
|
||||
|
||||
| Question | Answer | Product | Tags | Priority |
|
||||
|----------|--------|---------|------|----------|
|
||||
| How to return product? | Return possible within 14 days from receipt. Product must be in original packaging, with preserved appearance. Contact support to process return. | All | Return | 10 |
|
||||
| When will money be returned? | Refund processed within 5-10 business days after product received at our warehouse. | All | Return | 8 |
|
||||
|
||||
> 💡 **Recommendation**: Add minimum 20-30 questions for quality AI work. More questions = more accurate answers!
|
||||
|
||||
### 2.4 Activating Table as AI Source
|
||||
|
||||
1. In top right corner of table, find **⚙️ Table Settings**
|
||||
2. Enable toggle **"Use as AI Source"** ✅
|
||||
3. Click **"Save"**
|
||||
|
||||
> ✅ **Done!** Table is now indexed for vector search
|
||||
|
||||
---
|
||||
|
||||
## Step 3: Configuring AI Provider (Ollama)
|
||||
|
||||
### 3.1 Opening Ollama Settings
|
||||
|
||||
1. Go to **Settings** → **Integrations**
|
||||
2. Find **"Ollama"** block and click **"Details"**
|
||||
|
||||
### 3.2 Checking Base URL
|
||||
|
||||
1. Check **Base URL** field:
|
||||
- For Docker: `http://ollama:11434` ✅
|
||||
- For local: `http://localhost:11434`
|
||||
2. If URL is incorrect, fix and click **"Save"**
|
||||
|
||||
### 3.3 Selecting Model
|
||||
|
||||
1. In **"Model (LLM)"** field, select installed model:
|
||||
- `qwen2.5:7b` (recommended for Russian)
|
||||
2. In **"Embeddings Model"** field, select:
|
||||
- `mxbai-embed-large:latest`
|
||||
3. Click **"Save"**
|
||||
|
||||
---
|
||||
|
||||
## Step 4: Configuring AI Assistant
|
||||
|
||||
### 4.1 Opening Assistant Settings
|
||||
|
||||
1. Go to **Settings** → **Integrations**
|
||||
2. Find **"AI Assistant"** block and click **"Details"**
|
||||
|
||||
### 4.2 Configuring System Prompt
|
||||
|
||||
In **"System Prompt"** field, enter instructions for AI:
|
||||
|
||||
**Basic Prompt (to start)**:
|
||||
|
||||
```
|
||||
You are a professional customer support assistant.
|
||||
|
||||
Rules:
|
||||
1. Answer politely and professionally
|
||||
2. Use information from knowledge base
|
||||
3. If information not available — suggest contacting operator
|
||||
4. Answer briefly and to the point
|
||||
5. Always end with question "How else can I help?"
|
||||
```
|
||||
|
||||
**Advanced Prompt (with personalization)**:
|
||||
|
||||
```
|
||||
You are a professional customer support assistant for "Your Company Name".
|
||||
|
||||
About company:
|
||||
- We are engaged in [brief business description]
|
||||
- Our values: quality, reliability, customer focus
|
||||
|
||||
Communication style:
|
||||
- Friendly but professional
|
||||
- Address customer as "You"
|
||||
- Use emojis moderately (1-2 per message)
|
||||
|
||||
Response rules:
|
||||
1. MANDATORY: Answer ONLY in Russian. All questions and answers must be in Russian
|
||||
2. First search answer in knowledge base (RAG)
|
||||
3. If found — answer based on found information
|
||||
4. If not found — honestly say and suggest operator help
|
||||
5. Don't make up information about prices, terms, conditions
|
||||
6. For complex questions suggest contacting manager
|
||||
|
||||
Always end with: "How else can I help? 😊"
|
||||
```
|
||||
|
||||
### 4.3 Selecting Models
|
||||
|
||||
1. **LLM Model**: Select `qwen2.5:7b (ollama)`
|
||||
2. **Embedding Model**: Select `mxbai-embed-large:latest (ollama)`
|
||||
|
||||
> 💡 **Tip**: Models automatically pulled from Ollama settings
|
||||
|
||||
> 📊 **Context Window Size**:
|
||||
> - **Qwen2.5:7b**: Base context = **32,768 tokens** (~24,000 Russian words)
|
||||
> - Total data sent to model:
|
||||
> - System prompt: ~500-2000 characters (~300-1200 tokens)
|
||||
> - Conversation history: up to 20 messages (~100-500 tokens per message = ~2000-10000 tokens)
|
||||
> - RAG context: ~500-2000 tokens (from found data)
|
||||
> - Current question: ~50-200 tokens
|
||||
> - **Total**: approximately 3,000-15,000 tokens (sufficient reserve)
|
||||
> - If larger context needed → use Qwen3 with YaRN (up to 131K tokens)
|
||||
|
||||
### 4.4 Selecting RAG Table
|
||||
|
||||
1. In **"Selected RAG Tables"** field, select created table:
|
||||
- `FAQ - Frequently Asked Questions`
|
||||
2. Click **"Save"**
|
||||
|
||||
### 4.5 Configuring AI Rules
|
||||
|
||||
Create rule set for managing AI behavior:
|
||||
|
||||
1. Click **"Create"** button next to "Rule Set" field
|
||||
2. In modal window, fill in:
|
||||
|
||||
**Name**: `Hybrid Mode (RAG + Generation)`
|
||||
|
||||
**Description**: `AI first searches knowledge base, if not found — generates answer`
|
||||
|
||||
**Rules (JSON)**:
|
||||
```json
|
||||
{
|
||||
"checkUserTags": true,
|
||||
"searchRagFirst": true,
|
||||
"generateIfNoRag": true,
|
||||
"temperature": 0.7,
|
||||
"maxTokens": 500
|
||||
}
|
||||
```
|
||||
|
||||
3. Click **"Save"**
|
||||
4. Select created rule in dropdown
|
||||
|
||||
> 💡 **What parameters mean**:
|
||||
> - `checkUserTags: true` — consider user tags when searching
|
||||
> - `searchRagFirst: true` — first search knowledge base
|
||||
> - `generateIfNoRag: true` — generate answer if nothing found
|
||||
> - `temperature: 0.7` — balance between accuracy and creativity (0.0-1.0)
|
||||
> - `maxTokens: 500` — maximum answer length
|
||||
|
||||
### 4.6 RAG Search Settings
|
||||
|
||||
Expand **"🔍 RAG Search Settings"** section:
|
||||
|
||||
**Basic Settings:**
|
||||
1. **Search Method**: Select `Hybrid Search` (recommended)
|
||||
2. **Maximum Results**: `5`
|
||||
3. **Relevance Threshold**: `0.1` (from 0.01 to 1.0)
|
||||
|
||||
**Keyword Extraction:**
|
||||
1. ✅ **Enable Keyword Extraction**
|
||||
2. **Minimum Word Length**: `3`
|
||||
3. **Maximum Keywords**: `10`
|
||||
4. ✅ **Remove Stop Words**
|
||||
|
||||
**Search Weights (for hybrid):**
|
||||
1. **Semantic Search**: `70%` (accuracy)
|
||||
2. **Keyword Search**: `30%` (speed)
|
||||
|
||||
**Additional Settings:**
|
||||
1. ✅ **Fuzzy Search** (for typos)
|
||||
2. ✅ **Word Stemming** (finds different word forms)
|
||||
3. ☐ **Synonym Search** (currently disabled)
|
||||
|
||||
### 4.7 Saving Settings
|
||||
|
||||
Click **"Save"** button at bottom of form.
|
||||
|
||||
---
|
||||
|
||||
## Step 5: Testing AI Assistant
|
||||
|
||||
### 5.1 Using Built-in Tester
|
||||
|
||||
1. On AI assistant settings page, scroll down to **"🔍 System Monitoring"** block
|
||||
2. In **"🧠 RAG Functionality Test"** section:
|
||||
- Ensure table `FAQ - Frequently Asked Questions` is selected
|
||||
- Enter test question: `How to pay for order?`
|
||||
- Click **"Test RAG"**
|
||||
3. Observe process:
|
||||
- 🔍 Searching answer in knowledge base... (vector search)
|
||||
- 🤖 Generating answer with AI... (LLM generation)
|
||||
- ✅ Done!
|
||||
4. Check result:
|
||||
- Answer from your table should be displayed
|
||||
- Score (similarity score): closer to 0 is better
|
||||
|
||||
> 💡 **Good Score**: from -300 to 0 (answer found)
|
||||
> ⚠️ **Bad Score**: greater than 300 (answer not found, AI will make up its own)
|
||||
|
||||
### 5.2 Testing via Web Chat
|
||||
|
||||
1. Go to application main page
|
||||
2. Find **"💬 Chat with AI"** widget (usually bottom right)
|
||||
3. Click widget to open chat
|
||||
4. Enter question: `How much does delivery cost?`
|
||||
5. Check AI response
|
||||
|
||||
**Expected Result:**
|
||||
```
|
||||
🤖 AI Assistant:
|
||||
Free delivery for orders from 5,000₽.
|
||||
For orders less than 5,000₽ delivery cost is 300₽.
|
||||
|
||||
How else can I help? 😊
|
||||
```
|
||||
|
||||
### 5.3 Testing Different Scenarios
|
||||
|
||||
Try asking various questions:
|
||||
|
||||
**✅ Question from knowledge base:**
|
||||
```
|
||||
User: "How to return product?"
|
||||
AI: [Answer from FAQ table]
|
||||
```
|
||||
|
||||
**⚠️ Question NOT from knowledge base:**
|
||||
```
|
||||
User: "What's the weather today?"
|
||||
AI: "Sorry, I specialize in questions about our company and products.
|
||||
For weather questions, please contact specialized services.
|
||||
How else can I help?"
|
||||
```
|
||||
|
||||
**🎯 Question with typo:**
|
||||
```
|
||||
User: "How to pay for order?" (typo)
|
||||
AI: [Will find correct answer thanks to fuzzy search]
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
## Step 6: Advanced Capabilities (Optional)
|
||||
|
||||
### 6.1 Creating Supplier Table
|
||||
|
||||
#### "Supplier Database" Table Structure
|
||||
|
||||
1. Create new table: `Supplier Database`
|
||||
2. Add columns:
|
||||
|
||||
| Column | Type | Description |
|
||||
|--------|------|-------------|
|
||||
| Company Name | Text | Supplier name |
|
||||
| Product Category | Multiple Choice | Electronics, Furniture, Clothing, etc. |
|
||||
| Contact Person | Text | Manager name |
|
||||
| Email | Text | Email address |
|
||||
| Phone | Text | Contact phone |
|
||||
| Prices | Text | Price list (brief description) |
|
||||
| Payment Terms | Text | Deferred payment, prepayment, etc. |
|
||||
| Minimum Order | Number | Minimum order amount |
|
||||
| Delivery Time | Text | Delivery terms |
|
||||
| Rating | Number | Rating from 1 to 10 |
|
||||
| Notes | Text | Additional information |
|
||||
|
||||
3. Activate as AI source
|
||||
4. Fill with your supplier data
|
||||
|
||||
#### Prompt for AI Purchaser
|
||||
|
||||
Add to system prompt:
|
||||
|
||||
```
|
||||
ADDITIONALLY - Working with Suppliers:
|
||||
|
||||
When user asks about suppliers:
|
||||
1. Search in "Supplier Database" base
|
||||
2. Filter by product category
|
||||
3. Sort by rating and terms
|
||||
4. Provide TOP-3 recommendations
|
||||
|
||||
Response format:
|
||||
🏆 TOP-3 suppliers for "[category]":
|
||||
|
||||
1. [Name] ⭐ [Rating]/10
|
||||
📧 [Email] | 📞 [Phone]
|
||||
💰 Terms: [Payment Terms]
|
||||
🚚 Delivery: [Delivery Time]
|
||||
📦 Minimum: [Minimum Order]₽
|
||||
|
||||
2. ...
|
||||
3. ...
|
||||
|
||||
I recommend contacting [First Supplier Name] — best terms.
|
||||
```
|
||||
|
||||
### 6.2 Creating Employee Training Table
|
||||
|
||||
#### "Employee Knowledge Base" Table Structure
|
||||
|
||||
1. Create new table: `Employee Knowledge Base`
|
||||
2. Add columns:
|
||||
|
||||
| Column | Type | Description |
|
||||
|--------|------|-------------|
|
||||
| Question | Text | Employee question (purpose: Question for AI) |
|
||||
| Answer | Text | Detailed answer (purpose: AI Answer) |
|
||||
| Category | Multiple Choice | Sales, HR, Finance, IT, Marketing |
|
||||
| Department | Multiple Choice | Which department it's relevant for |
|
||||
| Difficulty | Number | From 1 (simple) to 5 (complex) |
|
||||
| Instructions | Text | Step-by-step instructions (if any) |
|
||||
| Links | Text | Links to documents/videos |
|
||||
| Update Date | Date | When information was updated |
|
||||
|
||||
3. Example questions:
|
||||
|
||||
**"Sales" Category:**
|
||||
- "How to process customer return?"
|
||||
- "What discounts can be given to regular customers?"
|
||||
- "How to work with corporate customers?"
|
||||
|
||||
**"HR" Category:**
|
||||
- "How to apply for vacation?"
|
||||
- "Where to contact for sick leave?"
|
||||
- "How does new employee onboarding work?"
|
||||
|
||||
**"IT" Category:**
|
||||
- "How to get access to corporate email?"
|
||||
- "What to do with VPN problems?"
|
||||
- "How to create support ticket?"
|
||||
|
||||
### 6.3 Creating Table Relationships
|
||||
|
||||
#### Example: "Customers" → "Orders" Relationship
|
||||
|
||||
1. Create **"Customers"** table:
|
||||
- Name, Email, Phone, Status (VIP/Regular)
|
||||
|
||||
2. Create **"Orders"** table:
|
||||
- Order Number, Date, Amount
|
||||
|
||||
3. In "Orders" table, add column:
|
||||
- **Name**: `Customer`
|
||||
- **Type**: `Relation`
|
||||
- **Related Table**: select `Customers`
|
||||
- **Display Field**: select `Name`
|
||||
|
||||
4. Add another column in "Orders":
|
||||
- **Name**: `Customer Email`
|
||||
- **Type**: `Lookup`
|
||||
- **Relation Through**: select `Customer` column
|
||||
- **Lookup Field**: select `Email`
|
||||
|
||||
**Result**: When selecting customer, their Email is automatically filled!
|
||||
|
||||
#### Using AI with Related Tables
|
||||
|
||||
AI automatically understands relationships and can answer questions:
|
||||
|
||||
```
|
||||
User: "Show all orders for customer Ivanov"
|
||||
AI: [Searches Orders table, filters by customer Ivanov]
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
## Step 7: Telegram and Email Integration (Optional)
|
||||
|
||||
### 7.1 Setting Up Telegram Bot
|
||||
|
||||
1. Go to **Settings** → **Integrations** → **Telegram**
|
||||
2. Create bot via [@BotFather](https://t.me/botfather) in Telegram:
|
||||
- Send `/newbot`
|
||||
- Choose bot name and username
|
||||
- Copy **Bot Token**
|
||||
3. In DLE settings, enter:
|
||||
- **Bot Token**: paste token from BotFather
|
||||
- **Bot Username**: your bot username (e.g., `@mycompany_bot`)
|
||||
4. Click **"Save"**
|
||||
5. In AI assistant settings, select this Telegram bot in **"Telegram Bot"** field
|
||||
|
||||
**Result**: AI will respond to Telegram messages!
|
||||
|
||||
### 7.2 Setting Up Email Integration
|
||||
|
||||
1. Go to **Settings** → **Integrations** → **Email**
|
||||
2. Fill IMAP settings (for receiving emails):
|
||||
- **IMAP Host**: `imap.gmail.com` (for Gmail)
|
||||
- **IMAP Port**: `993`
|
||||
- **IMAP User**: your email
|
||||
- **IMAP Password**: app password (not main password!)
|
||||
3. Fill SMTP settings (for sending emails):
|
||||
- **SMTP Host**: `smtp.gmail.com`
|
||||
- **SMTP Port**: `587`
|
||||
- **SMTP User**: your email
|
||||
- **SMTP Password**: app password
|
||||
- **From Email**: email for sending
|
||||
4. Click **"Test IMAP"** and **"Test SMTP"** to verify
|
||||
5. Click **"Save"**
|
||||
6. In AI assistant settings, select this Email in **"Email Contact"** field
|
||||
|
||||
> ⚠️ **Important for Gmail**: Create "App Password" in Google security settings
|
||||
|
||||
**Result**: AI will automatically respond to incoming emails!
|
||||
|
||||
---
|
||||
|
||||
## Step 8: Monitoring and Optimization
|
||||
|
||||
### 8.1 Checking Service Status
|
||||
|
||||
1. Go to **Settings** → **Integrations** → **AI Assistant**
|
||||
2. Scroll down to **"🔍 System Monitoring"**
|
||||
3. Click **"🔄 Refresh Status"**
|
||||
4. Check statuses:
|
||||
- 🟢 **Backend**: should be "Running"
|
||||
- 🟢 **Postgres**: should be "Running"
|
||||
- 🟢 **Ollama**: should show number of models
|
||||
- 🟢 **Vector Search**: should be "Running"
|
||||
|
||||
> ⚠️ If something is red (🔴) — see "Troubleshooting" section below
|
||||
|
||||
### 8.2 Analyzing Answer Quality
|
||||
|
||||
Regularly check AI answer quality:
|
||||
|
||||
1. **Score in RAG Tester**:
|
||||
- **-300 to 0** ✅ — excellent match
|
||||
- **0 to 300** ⚠️ — average match
|
||||
- **>300** ❌ — match not found
|
||||
|
||||
2. **If Score is Bad**:
|
||||
- Add more similar questions to table
|
||||
- Use different phrasings of same question
|
||||
- Increase relevance threshold (e.g., to 0.2)
|
||||
|
||||
### 8.3 Optimizing RAG Settings
|
||||
|
||||
Experiment with settings to improve results:
|
||||
|
||||
**For More Accurate Answers:**
|
||||
```
|
||||
Search Method: Semantic
|
||||
Relevance Threshold: 0.05 (lower = stricter)
|
||||
Weights: Semantic 100% / Keywords 0%
|
||||
```
|
||||
|
||||
**For Faster Answers:**
|
||||
```
|
||||
Search Method: Keyword Search
|
||||
Maximum Results: 3
|
||||
```
|
||||
|
||||
**For Balance (Recommended):**
|
||||
```
|
||||
Search Method: Hybrid
|
||||
Weights: Semantic 70% / Keywords 30%
|
||||
Relevance Threshold: 0.1
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
## ✅ AI Assistant Ready to Work!
|
||||
|
||||
### What You Now Have
|
||||
|
||||
✅ **Local AI Assistant** without cloud dependency
|
||||
✅ **FAQ Knowledge Base** for customer responses
|
||||
✅ **Vector Search** for accurate answers
|
||||
✅ **Configured Rules** for AI behavior
|
||||
✅ **Monitoring System** for quality control
|
||||
|
||||
### Economic Impact
|
||||
|
||||
With proper AI assistant setup, you will get:
|
||||
|
||||
✅ **Task Automation** - freeing time for strategy
|
||||
✅ **Improved Service Quality** - AI works 24/7 without fatigue
|
||||
✅ **Reduced Operating Costs** - less staff on routine tasks
|
||||
✅ **Faster Decision Making** - instant access to information
|
||||
|
||||
> 💡 **Detailed Information**: See [AI Assistant - Complete Description](./ai-assistant.md#economic-impact) - detailed description of all capabilities, use cases, and savings calculations.
|
||||
|
||||
---
|
||||
|
||||
## 📚 Next Steps
|
||||
|
||||
### Expand AI Capabilities
|
||||
|
||||
1. **Add More Tables**:
|
||||
- Partner knowledge base
|
||||
- Employee instructions
|
||||
- Product catalog
|
||||
- Contact database
|
||||
|
||||
2. **Create Rules for Different Scenarios**:
|
||||
- Strict mode (RAG only) — for finance
|
||||
- Creative mode (more generation) — for marketing
|
||||
- Hybrid mode (balance) — for support
|
||||
|
||||
3. **Integrate with Other Systems**:
|
||||
- CRM (customer synchronization)
|
||||
- Warehouse system (inventory)
|
||||
- Accounting (invoices and payments)
|
||||
|
||||
### Train Your Team
|
||||
|
||||
1. Show employees how AI works
|
||||
2. Explain how to add new questions to base
|
||||
3. Establish regular knowledge base update process
|
||||
4. Assign responsible person for AI answer quality
|
||||
|
||||
---
|
||||
|
||||
## 🆘 Troubleshooting
|
||||
|
||||
### Problem: Ollama Not Starting
|
||||
|
||||
**Symptoms**: Status "Ollama API not responding"
|
||||
|
||||
**Solution**:
|
||||
```bash
|
||||
# Check container
|
||||
docker ps | grep ollama
|
||||
|
||||
# Restart
|
||||
docker-compose restart ollama
|
||||
|
||||
# Check logs
|
||||
docker-compose logs ollama
|
||||
```
|
||||
|
||||
### Problem: AI Answers Inaccurately
|
||||
|
||||
**Symptoms**: Answers don't match knowledge base
|
||||
|
||||
**Solution**:
|
||||
1. Check Score in tester (should be < 300)
|
||||
2. Add more question variations to table
|
||||
3. Decrease relevance threshold (e.g., to 0.05)
|
||||
4. Check that columns have correct purposes ("Question for AI", "AI Answer")
|
||||
|
||||
### Problem: Vector Search Not Working
|
||||
|
||||
**Symptoms**: Vector Search status shows error
|
||||
|
||||
**Solution**:
|
||||
1. Check if Embedding model is installed
|
||||
2. Rebuild index: on table page click **"🔄 Rebuild Index"**
|
||||
3. Check that table is activated as AI source
|
||||
|
||||
### Problem: AI Answers in Wrong Language
|
||||
|
||||
**Symptoms**: Answers in English instead of Russian
|
||||
|
||||
**Solution**:
|
||||
1. Change system prompt, adding at start: `ALWAYS answer in Russian.`
|
||||
2. Use `qwen2.5:7b` model instead of `llama2:7b`
|
||||
3. In AI rules, set `"language": "ru"`
|
||||
|
||||
### Problem: Slow Responses
|
||||
|
||||
**Symptoms**: AI responds longer than 5-10 seconds
|
||||
|
||||
**Solution**:
|
||||
1. Use smaller model (`mistral:7b` instead of `qwen2.5:14b`)
|
||||
2. Decrease `maxResults` in RAG settings (e.g., to 3)
|
||||
3. Disable "Synonym Search" in additional settings
|
||||
4. Use SSD for storing models
|
||||
|
||||
---
|
||||
|
||||
## 📖 Additional Documentation
|
||||
|
||||
### Learn AI Capabilities
|
||||
|
||||
- 🤖 **[AI Assistant - Complete Description](./ai-assistant.md)** - all capabilities and use cases
|
||||
- 📊 **[Spreadsheet System](./tables-system.md)** - technical description of tables (temporary)
|
||||
- ⚙️ **[AI Configuration](./setup-ai-assistant.md#technical-documentation-for-developers)** - technical setup details
|
||||
|
||||
### General Documentation
|
||||
|
||||
- 🛡️ **[Security](./security.md)** - how AI protects your data
|
||||
- 💼 **[Blockchain for Business](./blockchain-for-business.md)** - AI integration with blockchain
|
||||
- 📋 **[FAQ](./FAQ.md)** - frequently asked questions
|
||||
|
||||
### Support
|
||||
|
||||
- 💬 **Support Chat**: https://hb3-accelerator.com/
|
||||
- 📧 **Email**: info@hb3-accelerator.com
|
||||
- 📚 **Knowledge Base**: https://hb3-accelerator.com
|
||||
|
||||
---
|
||||
|
||||
## 🔧 Technical Documentation (for Developers)
|
||||
|
||||
### AI System Architecture
|
||||
|
||||
```
|
||||
┌───────────────────────────────────────────────────────────┐
|
||||
│ AI Assistant Configuration in DLE │
|
||||
├───────────────────────────────────────────────────────────┤
|
||||
│ │
|
||||
│ 🤖 AI Providers: │
|
||||
│ ├── OpenAI (GPT-4, GPT-3.5) │
|
||||
│ ├── Anthropic (Claude) │
|
||||
│ ├── Google (Gemini) │
|
||||
│ └── Ollama (local models) │
|
||||
│ │
|
||||
│ ⚙️ AI Settings: │
|
||||
│ ├── System Prompt │
|
||||
│ ├── LLM Model Selection │
|
||||
│ ├── Embedding Model Selection │
|
||||
│ ├── RAG Table Selection │
|
||||
│ ├── Rules │
|
||||
│ └── RAG Search Settings │
|
||||
│ │
|
||||
│ 📋 Rules: │
|
||||
│ ├── JSON behavior configuration │
|
||||
│ ├── Create/edit/delete │
|
||||
│ └── Link to AI assistant │
|
||||
│ │
|
||||
│ 🔗 Integrations: │
|
||||
│ ├── Email (IMAP + SMTP) │
|
||||
│ └── Telegram Bot │
|
||||
│ │
|
||||
│ 🔍 RAG: │
|
||||
│ ├── Table selection for search │
|
||||
│ ├── Search settings (hybrid/semantic) │
|
||||
│ ├── Relevance threshold │
|
||||
│ └── Keyword extraction │
|
||||
│ │
|
||||
│ 📊 Monitoring: │
|
||||
│ ├── Service status (Backend, Ollama, Postgres) │
|
||||
│ ├── RAG functionality test │
|
||||
│ └── Progress tracking │
|
||||
│ │
|
||||
└───────────────────────────────────────────────────────────┘
|
||||
```
|
||||
|
||||
### Database
|
||||
|
||||
#### Table: `ai_providers_settings`
|
||||
|
||||
```sql
|
||||
CREATE TABLE IF NOT EXISTS ai_providers_settings (
|
||||
id SERIAL PRIMARY KEY,
|
||||
provider_encrypted TEXT, -- Provider: openai, anthropic, google, ollama
|
||||
api_key_encrypted TEXT, -- API key (encrypted)
|
||||
base_url_encrypted TEXT, -- Base URL for API
|
||||
selected_model_encrypted TEXT, -- Selected LLM model
|
||||
embedding_model_encrypted TEXT, -- Selected Embedding model
|
||||
created_at TIMESTAMP NOT NULL DEFAULT NOW(),
|
||||
updated_at TIMESTAMP NOT NULL DEFAULT NOW()
|
||||
);
|
||||
```
|
||||
|
||||
#### Table: `ai_assistant_settings`
|
||||
|
||||
```sql
|
||||
CREATE TABLE IF NOT EXISTS ai_assistant_settings (
|
||||
id SERIAL PRIMARY KEY,
|
||||
system_prompt_encrypted TEXT, -- System prompt
|
||||
selected_rag_tables INTEGER[], -- Array of RAG table IDs
|
||||
languages TEXT[], -- Array of supported languages
|
||||
model_encrypted TEXT, -- Selected LLM model
|
||||
embedding_model_encrypted TEXT, -- Selected Embedding model
|
||||
rules JSONB, -- Rules (DEPRECATED)
|
||||
rules_id INTEGER REFERENCES ai_assistant_rules(id), -- Rule reference
|
||||
telegram_settings_id INTEGER, -- Telegram bot reference
|
||||
email_settings_id INTEGER, -- Email settings reference
|
||||
rag_settings JSONB, -- RAG search settings
|
||||
updated_at TIMESTAMP DEFAULT NOW(),
|
||||
updated_by INTEGER
|
||||
);
|
||||
```
|
||||
|
||||
#### Table: `ai_assistant_rules`
|
||||
|
||||
```sql
|
||||
CREATE TABLE IF NOT EXISTS ai_assistant_rules (
|
||||
id SERIAL PRIMARY KEY,
|
||||
name TEXT NOT NULL, -- Rule set name
|
||||
description TEXT, -- Rule description
|
||||
rules JSONB NOT NULL, -- JSON configuration
|
||||
rules_encrypted TEXT, -- Encrypted version of rules
|
||||
created_at TIMESTAMP DEFAULT NOW(),
|
||||
updated_at TIMESTAMP DEFAULT NOW()
|
||||
);
|
||||
```
|
||||
|
||||
### Backend API
|
||||
|
||||
#### AI Provider Settings
|
||||
|
||||
- **GET** `/settings/ai-settings/:provider` — Get provider settings
|
||||
- **PUT** `/settings/ai-settings/:provider` — Save provider settings
|
||||
- **DELETE** `/settings/ai-settings/:provider` — Delete provider settings
|
||||
- **GET** `/settings/ai-settings/:provider/models` — Get model list
|
||||
- **POST** `/settings/ai-settings/:provider/verify` — Verify API key
|
||||
|
||||
#### AI Assistant Settings
|
||||
|
||||
- **GET** `/settings/ai-assistant` — Get assistant settings
|
||||
- **PUT** `/settings/ai-assistant` — Save assistant settings
|
||||
|
||||
#### AI Rules
|
||||
|
||||
- **GET** `/settings/ai-assistant-rules` — Get all rules
|
||||
- **GET** `/settings/ai-assistant-rules/:id` — Get rule by ID
|
||||
- **POST** `/settings/ai-assistant-rules` — Create rule
|
||||
- **PUT** `/settings/ai-assistant-rules/:id` — Update rule
|
||||
- **DELETE** `/settings/ai-assistant-rules/:id` — Delete rule
|
||||
|
||||
#### Ollama (Local Models)
|
||||
|
||||
- **GET** `/ollama/status` — Check Ollama status
|
||||
- **GET** `/ollama/models` — Get model list
|
||||
- **POST** `/ollama/install` — Install model
|
||||
- **DELETE** `/ollama/models/:modelName` — Delete model
|
||||
|
||||
### Frontend Pages
|
||||
|
||||
- **`/settings/ai`** — Main integrations page
|
||||
- **`/settings/ai/:provider`** — AI provider settings
|
||||
- **`/settings/ai/assistant`** — AI assistant settings
|
||||
|
||||
### Message Processing Flow
|
||||
|
||||
```
|
||||
1. User → Message
|
||||
↓
|
||||
2. UnifiedMessageProcessor
|
||||
↓
|
||||
3. Language Check (Russian only)
|
||||
↓
|
||||
4. Deduplication (message hash)
|
||||
↓
|
||||
5. Load Settings (aiAssistantSettingsService)
|
||||
↓
|
||||
6. Load Rules (aiAssistantRulesService)
|
||||
↓
|
||||
7. RAG Search (ragService)
|
||||
├── Semantic Search (vector search)
|
||||
├── Keyword Search
|
||||
└── Hybrid Search
|
||||
↓
|
||||
8. Generate Answer (generateLLMResponse)
|
||||
├── System Prompt
|
||||
├── Conversation History
|
||||
├── RAG Context
|
||||
└── Rules
|
||||
↓
|
||||
9. Answer → User
|
||||
```
|
||||
|
||||
### Security
|
||||
|
||||
- **Encryption**: All sensitive fields encrypted with AES-256
|
||||
- **Access Rights**: Only administrators can change settings
|
||||
- **Validation**: All input data and API keys validated
|
||||
|
||||
---
|
||||
|
||||
**© 2024-2025 Tarabanov Alexander Viktorovich. All rights reserved.**
|
||||
|
||||
**Document Version**: 1.0.0
|
||||
**Creation Date**: October 25, 2025
|
||||
|
||||
196
docs-en/setup-instruction.md
Normal file
196
docs-en/setup-instruction.md
Normal file
@@ -0,0 +1,196 @@
|
||||
<!--
|
||||
Copyright (c) 2024-2025 Tarabanov Alexander Viktorovich
|
||||
All rights reserved.
|
||||
|
||||
This software is proprietary and confidential.
|
||||
Unauthorized copying, modification, or distribution is prohibited.
|
||||
|
||||
For licensing inquiries: info@hb3-accelerator.com
|
||||
Website: https://hb3-accelerator.com
|
||||
GitHub: https://github.com/VC-HB3-Accelerator
|
||||
-->
|
||||
|
||||
# Digital Legal Entity Application Setup Instructions
|
||||
|
||||
## 🚀 Complete System Initialization Process
|
||||
|
||||
This document describes the complete process of preparing the application for work with blockchain support, smart contracts, and access management system.
|
||||
|
||||
---
|
||||
|
||||
## Step 1: Software Installation
|
||||
|
||||
1. Clone the project repository to your local device
|
||||
2. Run the application via Docker Compose or locally depending on configuration
|
||||
3. Open the web application in browser: `http://localhost:9000` (production) or `http://localhost:5173` (dev mode)
|
||||
|
||||
---
|
||||
|
||||
## Step 2: Crypto Wallet Connection
|
||||
|
||||
1. Make sure you have a browser wallet installed (MetaMask, WalletConnect, or similar)
|
||||
2. In the wallet, create or import an account with governance token
|
||||
3. In the web application, click the **"Connect Wallet"** button
|
||||
4. Select wallet type and confirm connection
|
||||
5. After successful connection, you will see your account address in the top corner
|
||||
|
||||
---
|
||||
|
||||
## Step 3: Adding RPC Providers (Security → RPC Providers)
|
||||
|
||||
1. Go to **Settings** → **Security** tab
|
||||
2. Find the **"RPC Providers"** section
|
||||
3. Click the **"Add"** button
|
||||
4. Fill out the form for each blockchain network you want to use:
|
||||
- **Network Name** (e.g.: Ethereum, Polygon, BSC)
|
||||
- **RPC URL** (connection link, example: `https://eth-mainnet.g.alchemy.com/v2/YOUR-API-KEY`)
|
||||
- **Network ID** (Chain ID)
|
||||
5. Click **"Save"** for each provider
|
||||
6. The system will automatically verify connection correctness
|
||||
|
||||
> ⚠️ **Important**: Get API keys from providers (Alchemy, Infura, Quicknode, etc.) before adding
|
||||
|
||||
---
|
||||
|
||||
## Step 4: Multi-Deploy Smart Contract Setup
|
||||
|
||||
1. Go to **Settings** → **Blockchain** tab
|
||||
2. Fill out the form
|
||||
3. Click **"Launch Deploy"**
|
||||
|
||||
---
|
||||
|
||||
## Step 5: Deploy Completion and Contract Address Saving
|
||||
|
||||
1. Wait for deploy completion (depends on network, usually 30-120 seconds)
|
||||
2. After successful completion, the **"Contract Management"** page will open
|
||||
3. **Copy the deployed contract address** (usually looks like: `0x742d35Cc6634C0532925a3b844Bc...`)
|
||||
|
||||
---
|
||||
|
||||
## Step 6: Smart Contract Authentication Setup
|
||||
|
||||
1. Return to **Settings** → **Authentication** tab
|
||||
2. In the **"Smart Contract Address"** field, paste the address copied in step 5
|
||||
3. Set thresholds for access management:
|
||||
- **Minimum number of tokens for editing** (e.g.: 100 tokens)
|
||||
- **Minimum number of tokens for viewing** (e.g.: 1 token)
|
||||
|
||||
---
|
||||
|
||||
## Step 7: AI and Database Setup
|
||||
|
||||
1. Go to **Settings** → **AI** tab
|
||||
2. Open the **"Database"** subsection
|
||||
3. Replace default passwords
|
||||
4. Click **"Generate New Encryption Key"**
|
||||
- The system will automatically create a cryptographic key
|
||||
- **Save the key in a secure place** (it will be needed for data recovery)
|
||||
|
||||
---
|
||||
|
||||
## Step 8: Internet Access Setup (Optional)
|
||||
|
||||
**If you need access to the web application from outside via internet:**
|
||||
|
||||
1. Go to **Settings** → **Server** tab
|
||||
2. On the **Server** page, select **WEB SSH** or another suitable service
|
||||
3. Fill out the form to migrate local application to a virtual device with:
|
||||
- **Public IP address**
|
||||
- **Connection to your domain name**
|
||||
4. Click **"Publish"**
|
||||
5. Wait for migration process completion
|
||||
|
||||
> ℹ️ **Note**: This step requires having a registered domain name and access to DNS settings
|
||||
|
||||
---
|
||||
|
||||
## Step 9: Legal Documents Setup for Personal Data Processing
|
||||
|
||||
### 9.1 Filling Company Legal Information
|
||||
|
||||
1. Go to **CRM** → **Content** section
|
||||
2. Find and open the **"Company Legal Information"** form
|
||||
3. Fill in all necessary fields:
|
||||
- **Full Organization Name** (legal name)
|
||||
- **Short Name**
|
||||
- **Legal Form** (LLC, sole proprietor, JSC, etc.)
|
||||
- **Legal Address**
|
||||
- **Actual Address** (if different)
|
||||
- **INN / OGRN / KPP** (registration data)
|
||||
- **Contact Information** (phone, email, website)
|
||||
- **Responsible Person for Personal Data Processing** (Full Name, position)
|
||||
- **Applicable Jurisdiction** (GDPR, CCPA, Russian legislation, etc.)
|
||||
4. Click **"Save"**
|
||||
|
||||
> 💡 **Hint**: All entered data will automatically be inserted into all legal document templates
|
||||
|
||||
### 9.2 Working with Document Templates
|
||||
|
||||
1. In the **Content** section, go to the **"Templates"** subsection
|
||||
2. Select necessary document templates required by regulators:
|
||||
- **Privacy Policy**
|
||||
- **User Agreement**
|
||||
- **Personal Data Processing Consent**
|
||||
- **Cookie Usage Policy**
|
||||
3. For each template:
|
||||
- Click **"Preview"** to check automatically filled data
|
||||
- Edit specific data processing parameters if necessary
|
||||
- Select action:
|
||||
- **"Publish for Public Use"** — document will be available on the site
|
||||
- **"Publish for Internal Use"** — document available only within CRM
|
||||
- **"Print"** — export to PDF for printing or signing
|
||||
4. Confirm publication
|
||||
5. The system will automatically add documents to corresponding application pages
|
||||
|
||||
> ⚠️ **Important**: It is recommended to consult with a lawyer before publishing documents to ensure full compliance with legal requirements
|
||||
|
||||
---
|
||||
|
||||
## ✅ Application Ready to Work!
|
||||
|
||||
After completing all steps, your application is fully configured and ready to use.
|
||||
|
||||
**Next Steps:**
|
||||
- 📖 AI Assistant Setup (see document: `setup-ai-assistant.md`)
|
||||
- 🔐 Smart Contract Management (see document: `manage-smart-contracts.md`)
|
||||
|
||||
---
|
||||
|
||||
## 🆘 Security Recommendations
|
||||
|
||||
✓ Keep contract addresses and encryption keys in a secure place
|
||||
✓ Use strong passwords for DB
|
||||
✓ Regularly create configuration backups
|
||||
✓ Never share wallet private keys
|
||||
✓ Use HTTPS for application access in production
|
||||
|
||||
---
|
||||
|
||||
## 📝 What's Next?
|
||||
|
||||
After completing basic setup, you can:
|
||||
1. Add users and manage their permissions
|
||||
2. Create groups for collaboration
|
||||
3. Configure AI assistant for task automation
|
||||
4. Manage smart contracts to extend functionality
|
||||
5. Integrate external services and bots
|
||||
|
||||
---
|
||||
|
||||
## 📚 Additional Documentation
|
||||
|
||||
### Explore DLE Capabilities
|
||||
- 🤖 **[AI Assistant](./ai-assistant.md)** - learn how AI can become a second pilot for your team
|
||||
- 💼 **[Blockchain for Business](./blockchain-for-business.md)** - how asset tokenization solves business tasks
|
||||
- 🛡️ **[Security](./security.md)** - multi-level protection for your business
|
||||
|
||||
### Technical Information
|
||||
- 🔗 **[Blockchain Technical Documentation](./blockchain-integration-technical.md)** - for developers
|
||||
- 📋 **[FAQ](./FAQ.md)** - frequently asked questions
|
||||
- 📝 **[Application Description](./application-description.md)** - functionality overview
|
||||
|
||||
### Support
|
||||
- 💬 **Support Chat**: https://hb3-accelerator.com/
|
||||
- 📧 **Email**: info@hb3-accelerator.com
|
||||
|
||||
132
docs-en/system-messages-management.md
Normal file
132
docs-en/system-messages-management.md
Normal file
@@ -0,0 +1,132 @@
|
||||
# Technical Specification: System Messages Management
|
||||
|
||||
## 1. Goal and Context
|
||||
- Provide managed display of system messages on the main page (`/`, component `HomeView.vue`) and add an administrative interface for their creation and moderation in the content section (`/content`, component `ContentListView.vue`).
|
||||
- System messages must support "draft" and "published" statuses, be stored in the database, and be accessible via REST API.
|
||||
|
||||
## 2. Current State
|
||||
- The main page is built by component `HomeView.vue` and displays the assistant chat (`ChatInterface.vue`), in which system messages (`Message.vue`) are highlighted by the `message.role === 'system'` attribute.
|
||||
- The content section (`ContentListView.vue`) contains navigation cards: "Create Page", "Templates", "Public", "Settings", "Internal". Cards lead to existing routes `content-create`, `content-templates`, `content-published`, `content-settings`, `content-internal`.
|
||||
- The project lacks entities and API for system messages; current `pagesService.js` works only with pages (`/pages`).
|
||||
|
||||
## 3. New User Scenarios
|
||||
- **Viewing System Messages (main page, `/`):**
|
||||
- Published system messages are loaded into the assistant chat and displayed as collapsed cards with clickable headers.
|
||||
- When clicking the header, the message expands: the chat feed displays the full message text **or** sends a pre-prepared response from the AI assistant (the "response" content is stored with the message and selected by the `reply_type` flag).
|
||||
- Messages must be explicitly marked as system (color, icon). When reopened, the user sees the last expansion state; local "read" memory is possible.
|
||||
- **"System Messages" Section (`/content`):**
|
||||
- On the `/content` page, a new "System Messages" card appears with a "Details" button. Navigation leads to a page with a user table (`/content/system-messages/table`), built on existing table components (see `UserTablesList.vue`), without a separate dashboard of cards.
|
||||
- The table displays system messages row by row, with multiple selection via checkboxes; available bulk actions: publish, unpublish, move to drafts, delete.
|
||||
- For each message, clicking "Details" (within the row) opens view/edit with a form (see below).
|
||||
- **Create/Edit (`/content/system-messages/create`, `/content/system-messages/:id/edit`):**
|
||||
- Form with fields: title, brief description, main text (Markdown/HTML), response type (`inline` — show content, `assistant_reply` — send prepared assistant response), "Assistant Reply" field (active when `assistant_reply`), importance tag (info/warning/danger), publication start date (optional), end date (optional), flag for guest display.
|
||||
- Buttons: "Save as Draft", "Publish". When editing — "Update", "Unpublish", "Delete".
|
||||
- Validations: title and main text required (or assistant reply in corresponding mode); date validation (end ≥ start).
|
||||
- **Working with System Messages Table:**
|
||||
- Columns: selection checkbox, title (clickable), status, response type, active period, target audience (guests/authenticated/all), creation date, author.
|
||||
- Bulk actions are performed for selected rows; single actions available via context menu/buttons in row (edit, publish, unpublish, delete).
|
||||
|
||||
## 4. Interface Requirements
|
||||
- In `ContentListView.vue`, add "System Messages" card to the `management-blocks` grid with a `Details` button. The card design should match existing blocks (header, description, button).
|
||||
- System messages table page:
|
||||
- Use `BaseLayout` and local styles (`scoped`).
|
||||
- Table supports sorting, filtering by status, and search by title.
|
||||
- Checkboxes in header and rows for bulk selection; action panel appears when selection exists.
|
||||
- "Create Message" button opens creation form.
|
||||
- Create/Edit Form:
|
||||
- Rich-text (minimum Markdown) with preview and character/word counters.
|
||||
- Display mode toggle (`inline`/`assistant_reply`) with conditional display of "Assistant Reply" field (can use `<transition>`).
|
||||
- Field for icon/color selection by `severity` (static presets).
|
||||
- Main Page:
|
||||
- System messages are displayed in the chat block as collapsed cards (`system-message-collapsed`). Clicking the header expands the card (`system-message-expanded`) or initiates assistant sending (UI shows "message from assistant").
|
||||
- For expanded messages, provide "Collapse" button and (optionally) "Mark as Read". Store state in `localStorage`.
|
||||
|
||||
## 5. Routing and Components
|
||||
- Add routes in `router/index.js`:
|
||||
- - `/content/system-messages/table` → `SystemMessagesTableView.vue`
|
||||
- - `/content/system-messages/create` → `SystemMessageCreateView.vue`
|
||||
- - `/content/system-messages/:id` → `SystemMessageDetailsView.vue` (view)
|
||||
- - `/content/system-messages/:id/edit` → `SystemMessageEditView.vue`
|
||||
- If needed for modal/nested routes, child routes or named views can be used.
|
||||
- Create corresponding Vue components in `src/views/content/system-messages/` and a common set of reusable elements (table, form, filters, bulk actions) in `src/components/system-messages/`.
|
||||
- Create service `src/services/systemMessagesService.js` with methods for the new API.
|
||||
|
||||
## 6. API and Data Requirements
|
||||
- **New Table** `system_messages` (PostgreSQL):
|
||||
- `id` (uuid, pk)
|
||||
- `title` (text, not null)
|
||||
- `summary` (text, nullable)
|
||||
- `content` (text, not null)
|
||||
- `reply_type` (enum: inline, assistant_reply; default inline)
|
||||
- `assistant_reply_content` (text, nullable; required when `reply_type = assistant_reply`)
|
||||
- `severity` (enum: info, warning, danger; default info)
|
||||
- `status` (enum: draft, published; not null)
|
||||
- `visible_for` (enum: all, authenticated, guests; default all)
|
||||
- `publish_at` (timestamp, nullable)
|
||||
- `expire_at` (timestamp, nullable)
|
||||
- `created_at`, `updated_at`
|
||||
- `created_by`, `updated_by` (references users/identities, nullable)
|
||||
- `slug` (text, unique, for addressing by link if needed)
|
||||
- **REST API (Express):**
|
||||
- `GET /system-messages` (pagination, filters by status, search)
|
||||
- `GET /system-messages/published` (filtering by date/audience; public)
|
||||
- `GET /system-messages/:id` (access only for authorized editors)
|
||||
- `POST /system-messages` (creation; `MANAGE_LEGAL_DOCS` permission)
|
||||
- `PATCH /system-messages/:id` (editing; status checks)
|
||||
- `DELETE /system-messages/:id` (soft delete or physical)
|
||||
- `POST /system-messages/:id/publish` and `POST /system-messages/:id/unpublish` (optional, if not using PATCH)
|
||||
- All protected endpoints must require authorization and permissions (see `permissions.js`, `usePermissions`).
|
||||
- Add new migration (`backend/scripts/run-migrations.js`) and ORM/SQL files in the project's existing format.
|
||||
- Update logging and error handling `winston`, add input validation (e.g., `Joi` or custom).
|
||||
|
||||
## 7. Frontend Display Logic
|
||||
- `HomeView.vue`:
|
||||
- On initialization, request published system messages (considering current audience) via `systemMessagesService.getPublished({ includeExpired: false })`.
|
||||
- Cache response in store or local state; when subscribing to WebSocket, can provide `system_message_updated` event.
|
||||
- Add expansion handler: on header click, either substitute full message text (`inline`), or initiate sending `assistant_reply_content` to chat (without user participation).
|
||||
- Add message hiding handler, saving identifier in `localStorage` and filtering locally.
|
||||
- `ContentListView.vue`:
|
||||
- Add new "System Messages" card to the `management-blocks` grid, without breaking the adaptive grid (update `grid-template-columns` if needed).
|
||||
- List Pages:
|
||||
- Implement pagination (lazy loading or regular), sorting by date.
|
||||
- For statuses, use color badges (info/warning/danger).
|
||||
- Creation Form:
|
||||
- Support submit via `yarn lint`-friendly code; client-side validation (e.g., using `computed`/`watch`).
|
||||
- On successful publication, redirect to published list; when saving draft — stay on page with notification.
|
||||
|
||||
## 8. Security and Access Requirements
|
||||
- Creation/editing scenarios available only to roles with `PERMISSIONS.MANAGE_LEGAL_DOCS`.
|
||||
- Public list (`GET /system-messages/published`) filters by:
|
||||
- `status === 'published'`.
|
||||
- `publish_at <= now()` (or null).
|
||||
- `expire_at > now()` (or null).
|
||||
- `visible_for` is checked based on context (guest/authenticated).
|
||||
- When issuing through chat, hide fields `created_by`, `updated_by`, internal tags.
|
||||
- Consider CSRF, CORS, rate-limit (adopt config from existing routes).
|
||||
|
||||
## 9. Testing
|
||||
- **Backend:**
|
||||
- Unit tests for CRUD in `tests/system-messages/*.test.js` (Mocha).
|
||||
- Check publish/expire filters and role-based access.
|
||||
- Test migration (rollback/apply).
|
||||
- **Frontend:**
|
||||
- Vue unit tests (if configured) for main components (form, list).
|
||||
- E2E (if available) — scenario: create draft → publish → display on main page.
|
||||
- **Regression Checks:**
|
||||
- Ensure existing content list and assistant chat continue to work without errors (`yarn lint`, `yarn test`).
|
||||
|
||||
## 10. Integration and DevOps
|
||||
- Update `docker-compose.yml` if needed (e.g., add migrations to startup process).
|
||||
- Ensure new environment variables (if any, e.g., message count limits) are documented in `README.md` and `setup-instruction.md`.
|
||||
- Add seeding script (optional) for test system messages.
|
||||
|
||||
## 11. Open Questions
|
||||
- Is publication history (auditing) needed? If yes — provide `system_messages_history` table.
|
||||
- Is multilingual support required? (If absent — limit to one language, EN).
|
||||
- Is WebSocket notification needed when new messages appear? (If yes — add event to `wsHub.js`).
|
||||
|
||||
## 12. Final Artifacts
|
||||
- Backend: new routes, controllers, service, migration.
|
||||
- Frontend: new pages and service, updated routes and components `HomeView`, `ContentListView`.
|
||||
- Documentation: update `README.md` (launch section), `application-description.md` or `tables-system.md` when schemas change, this specification.
|
||||
|
||||
1639
docs-en/tables-system.md
Normal file
1639
docs-en/tables-system.md
Normal file
File diff suppressed because it is too large
Load Diff
Reference in New Issue
Block a user