🛠️ AI Service Desk Copilot

Agentic AI for enterprise IT: 80% automated ticket resolution via RAG + tools + DevOps pipeline. LangChain/LangGraph | Python | Docker.

🚀 Overview

Autonomous workflow:

• Perceive: Docs → ChromaDB RAG
• Reason: LLM planning (Groq)
• Act: APIs (Jira/Email)
• Deploy: GitHub Actions → Cloud

Live Demo: [Post-Stage 6]

📋 User Stories

As IT Agent:

1. Ingest docs → Vector DB
2. Analyze ticket → Match solutions
3. Auto-resolve (APIs)
4. Escalate if needed
5. Log for audits

As DevOps: 6. CI/CD deploy

🗺️ Architecture

graph TD
    A[Ticket Input] --> B[Retriever Agent<br/>RAG Query]
    B --> C[Planner Agent<br/>LangGraph]
    C --> D{Confidence >80%?}
    D -->|Yes| E[Executor Agent<br/>API Tools]
    D -->|No| F[Human Escalate]
    E --> G[Resolution Email]
    G --> H[Monitor & Reflect]
    F --> H
    style C fill:#f0f8ff

Loading