Project LoopBack

LoopBack is an AI-powered IT Support Helpdesk system that acts as a first line of defense for support teams. It intelligently handles user inquiries using a Knowledge Base (KB) and escalates complex issues to human agents when necessary. Crucially, it learns from every resolved ticket to improve its future responses.

Key Features

• Intelligent Chat Interface: Users converse naturally with the AI to troubleshoot issues.
• Automatic Escalation: If the AI cannot resolve an issue (or if hardware/admin intervention is required), it automatically drafts a ticket with a summary of the problem and the full conversation history.
• Knowledge Base Integration:
  • Retrieval: Uses fuzzy search logic to find relevant solutions from a CSV database (knowledge_base/Workplace_IT_Support_Database.csv).
  • Robust Search: Matches against "Issue", "Question", and "Tags", ignoring punctuation and case.
  • Duplicate Prevention: Automatically blocks duplicate or highly similar questions from being added to the KB to keep it clean.
• Self-Learning: When an admin marks a ticket as "Resolved" with a quality answer, the system automatically adds that solution to the Knowledge Base for future use.
• Admin Dashboard: View and manage tickets, see AI-drafted solutions, and monitor KB updates.
• Multi-Channel Support:
  • Discord Bot: Users can open tickets directly from Discord. The bot creates dedicated threads (private/public) for each issue to keep channels clean.
  • Web Portal: A responsive React application for tracking and managing tickets.

Technology Stack

• Backend: Python (FastAPI)
• Frontend: React (Vite + Tailwind CSS + Lucide Icons)
• AI Model: Google Gemini-3-Pro
• Database: JSON file (tickets_db.json) for tickets, CSV file for Knowledge Base.
• Integration: Discord.py (Bot)

System Architecture

graph TD
    User([User]) <--> Frontend[React Frontend]
    User([User]) <--> Discord[Discord Bot]
    Frontend <--> API[FastAPI Backend]
    Discord <--> API
    
    subgraph Backend Services
        API <--> AI[Google Gemini AI]
        API <--> KB[(Knowledge Base CSV)]
        API <--> DB[(Tickets DB JSON)]
    end
    
    Admin([Admin]) <--> Frontend

Loading

Workflow

sequenceDiagram
    participant U as User
    participant F as Frontend
    participant B as Backend
    participant AI as Gemini AI
    participant K as Knowledge Base
    participant A as Admin

    U->>F: Asks Question
    U->>Discord: Asks Question
    F->>B: Sends Query
    Discord ->> B: Sends Query
    B->>K: Search for existing solutions
    K-->>B: Returns context
    B->>AI: Analyze query + Context
    AI-->>B: Returns Draft Response & Metadata
    
    alt Solution Found
        B-->>F: Returns AI Solution
        F-->>U: Displays Solution
    else Issue Unresolved
        B->>B: Creates Ticket (Status: Pending)
        B-->>F: Returns Ticket Created
        F-->>U: Notify Ticket Created
    end
    
    opt Admin Resolution
        A->>F: Reviews Ticket
        A->>B: Submits Final Answer
        B->>AI: Standardize Resolution
        AI-->>B: Returns Cleaned Text
        B->>K: Updates Knowledge Base
    end

Loading

Setup Instructions

Prerequisites

• Python 3.9+
• Node.js & npm
• Google Gemini API Key

Backend Setup

1. Navigate to the project root:

   cd LoopBack

2. Install dependencies:

   pip install -r requirements.txt

3. Set up your environment variables:

   • Create a .env file in the root directory.

   • GOOGLE_API_KEY=your_api_key_here

   • DISCORD_BOT_TOKEN=<bot-token>

   • DISCORD_GUILD_ID=<server-id> (The server ID where you want the bot to operate)

   • DISCORD_CHANNEL_ID=<channel-id> (The channel ID where you want the bot to operate)

   • If you wish to use Langsmith services, add LANGSMITH_TRACING=true LANGSMITH_ENDPOINT=https://api.smith.langchain.com LANGSMITH_API_KEY=<your-api-key> LANGSMITH_PROJECT=<your-project-name>

   • If you wish to use Langsmith services, add LANGSMITH_TRACING=true LANGSMITH_ENDPOINT=https://api.smith.langchain.com LANGSMITH_API_KEY=<your-api-key> LANGSMITH_PROJECT=<your-project-name>

   • Else, simply add the line LANGSMITH_TRACING=true

4. Start the backend server:

   python3 server.py

   The server will run on http://localhost:8000.

Frontend Setup

1. Navigate to the frontend directory:

   cd frontend

2. Install dependencies:

   npm install

3. Start the development server:

   npm run dev

   The application will be available at http://localhost:5173.
4. Run the discord bot

   python3 discord_bot.py

Usage

1. User Portal: Users open the app and type their IT issue. The AI attempts to solve it using the Knowledge Base.
2. Ticket Creation: If unresolved, a ticket is created.
3. Admin Resolution: An admin reviews the ticket via the dashboard (or simulates resolution via API) and provides a final answer.
4. Learning: The system detects the high-quality resolution and adds it to the Knowledge Base for next time.

Project Structure

• server.py: Main backend logic (App, API endpoints, AI integration).
• discord_bot.py: Discord bot logic.
• tickets_db.json: Stores all ticket data.
• knowledge_base/: Contains the CSV database used for RAG (Retrieval-Augmented Generation).
• frontend/: React source code.
  • src/UserPortal.jsx: The chat interface for end-users.
  • src/AdminDashboard.jsx: Interface for support agents.