Skip to content

Create Dobby AI Assistant with Python Neural Network Backend and Node.js Web Frontend#2

Draft
Copilot wants to merge 4 commits intomainfrom
copilot/create-ai-neural-network
Draft

Create Dobby AI Assistant with Python Neural Network Backend and Node.js Web Frontend#2
Copilot wants to merge 4 commits intomainfrom
copilot/create-ai-neural-network

Conversation

Copy link

Copilot AI commented Oct 13, 2025

🤖 Dobby AI Assistant - Complete Personal AI with Neural Network Memory

Implemented a full-featured personal AI assistant that combines Python's AI/ML capabilities with Node.js's web server efficiency. The system provides neural network-based semantic memory, privacy-focused internet search, and a beautiful web interface - all while keeping your data completely private on your local machine.

🏗️ Hybrid Architecture

The implementation uses both Python and Node.js strategically:

Python Backend (AI/ML Processing):

  • Neural network embeddings with sentence-transformers
  • Semantic memory search using vector similarity
  • Local SQLite database for privacy
  • Internet search integration (DuckDuckGo)

Node.js Frontend (Web Server):

  • Express API server
  • Real-time web interface
  • HTTP request handling
  • Bridge between browser and Python AI

This architecture leverages Python's strength in AI/ML libraries (transformers, torch, numpy) while using Node.js for what it does best - fast, efficient web servers and real-time user interaction.

✨ Key Features

  • 🧠 Neural Network Memory: Uses sentence-transformers to create semantic embeddings, enabling intelligent memory search based on meaning rather than just keywords
  • 🔒 Privacy-First: All data stored locally in SQLite - no external API calls required
  • 🌐 Internet Access: Integrated DuckDuckGo search for private web queries
  • 💬 Conversational Interface: Beautiful, responsive web UI with real-time responses
  • 📚 Context-Aware: Remembers conversations and learns from interactions
  • ⚡ Graceful Degradation: Works with basic keyword matching if ML libraries aren't installed

📸 Screenshots

Initial Interface:
Dobby AI Assistant Interface

Memory Recall in Action:
AI remembering user preferences

🚀 Usage

# Install dependencies
npm install
pip3 install -r python/requirements.txt

# Start the server
npm start

# Open http://localhost:3000

Tell Dobby things:

"Remember that I love Python for AI projects"
"I prefer Node.js for web development"

Ask questions:

"What are my programming preferences?"
"What is machine learning?"  // triggers internet search

📂 Project Structure

python/
  ├── memory_system.py      # Neural network memory with embeddings
  ├── search_system.py      # Privacy-focused internet search
  ├── dobby_ai.py          # Main AI processing core
  └── requirements.txt      # Python dependencies

public/
  └── index.html           # Interactive web interface

server.js                  # Express API server (Node.js)
package.json              # Node.js dependencies

📚 Documentation

  • README.md - Complete setup and usage documentation
  • ARCHITECTURE.md - Detailed explanation of Python vs Node.js architectural choices (Russian)
  • QUICKSTART.md - 5-minute quick start guide

🔐 Privacy & Security

  • ✅ All conversations stored locally
  • ✅ All memories in local SQLite database
  • ✅ Internet searches via DuckDuckGo (no tracking)
  • ✅ No external API calls (optional ML models run locally)
  • ✅ Complete control over your data

💡 Why This Architecture?

Python for AI: When you need neural networks, embeddings, semantic analysis, and machine learning - Python's ecosystem (transformers, torch, sentence-transformers) is unmatched.

Node.js for Web: When you need a fast web server, REST API, and real-time user interaction - Node.js's async architecture and JavaScript integration excels.

This hybrid approach delivers the best of both worlds, as detailed in ARCHITECTURE.md.

🛠️ Technology Stack

AI/ML: sentence-transformers, torch, transformers, numpy
Web: Express.js, Node.js, vanilla JavaScript
Storage: SQLite (local, private)
Search: DuckDuckGo, BeautifulSoup


This implementation provides a solid foundation for a personal AI assistant that respects privacy while delivering intelligent, context-aware responses with neural network-powered memory.

Warning

Firewall rules blocked me from connecting to one or more addresses (expand for details)

I tried to connect to the following addresses, but was blocked by firewall rules:

  • html.duckduckgo.com
    • Triggering command: python3 /home/REDACTED/work/Dobby-v01/Dobby-v01/python/dobby_ai.py query what is artificial intelligence? (dns block)

If you need me to access, download, or install something from one of these locations, you can either:

Original prompt

Хочу ддля начала создать у себя на меке ИИ который бы вел себя как нейро сеть , анализировала и запоминала бы все и могла подсказывать с выходом в интернет но нонфиденциальностью моих данных и базы ИИ подсказывай где нужен пайтон а где мне ближе - nodeJs & javascript


💬 Share your feedback on Copilot coding agent for the chance to win a $200 gift card! Click here to start the survey.

Copilot AI and others added 3 commits October 13, 2025 20:37
…s frontend

Co-authored-by: getmanele <25000832+getmanele@users.noreply.github.com>
Co-authored-by: getmanele <25000832+getmanele@users.noreply.github.com>
Co-authored-by: getmanele <25000832+getmanele@users.noreply.github.com>
Copilot AI changed the title [WIP] Add AI neural network for learning and suggestions Create Dobby AI Assistant with Python Neural Network Backend and Node.js Web Frontend Oct 13, 2025
Copilot AI requested a review from getmanele October 13, 2025 20:47
@getmanele
Copy link
Owner

ok

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants