🔧 NetFix AI

AI-Powered Network Troubleshooting Assistant for Telecom R&D Labs

Conversational AI agent that diagnoses network failures, identifies root causes, and recommends remediation — in under 60 seconds.

---

Overview

NetFix AI is a RAG-based network operations assistant built for telecom R&D lab environments. Engineers can query the system in natural language to get evidence-backed root cause analysis, remediation steps, and historical incident correlation across multiple lab networks.

Built for: Major US Telecom Operator | R&D Network Lab Division
Domain: Generative AI × Network Operations × Telecom Engineering

---

Features

Feature	Description
Natural Language RCA	Ask questions in plain English, get structured root cause analysis
Multi-Source Ingestion	Syslogs, SNMP metrics, device inventory, topology, incident tickets
Evidence Citation	Every claim cites exact log lines, timestamps, and metric values
Historical Correlation	Matches current symptoms against past incidents automatically
Interactive Topology Map	Live network graph with color-coded device status
Predictive Alerting	Detects metric trends before they hit critical thresholds
Knowledge Base Builder	Save resolved incidents for future correlation
Runbook Generator	Auto-generates downloadable PDF runbooks for any incident
Multi-turn Conversation	Full session memory with chat history
NOC Dashboard	Real-time network operations center view

---

Tech Stack

Layer	Technology
LLM	Groq (Llama 3.3 70B) — free tier
Embeddings	sentence-transformers (all-MiniLM-L6-v2) — local
Vector DB	ChromaDB — local persistent
Metrics Store	SQLite + Pandas
Orchestration	LangChain
Topology Viz	PyVis
UI	Streamlit
PDF Generation	ReportLab

---

Project Structure

NetFixAI/
├── app.py                      # NOC Dashboard (main landing page)
├── pages/
│   └── chat.py                 # AI Chat Interface
├── src/
│   ├── ingest.py               # Data ingestion pipeline
│   ├── tools.py                # semantic_search + metrics_query tools
│   ├── agent.py                # Groq LLM agent + conversation memory
│   ├── topology_viz.py         # PyVis interactive topology map
│   ├── predictor.py            # Predictive failure alerting
│   ├── knowledge_base.py       # KB save/search/retrieve
│   └── runbook_generator.py    # PDF runbook generator
├── data/
│   ├── router_syslog.log       # Syslog events
│   ├── device_inventory.csv    # Device metadata
│   ├── network_topology.json   # Network links and BGP peers
│   ├── snmp_metrics.csv        # SNMP time-series metrics
│   ├── incident_tickets.json   # Historical incident tickets
│   └── knowledge_base.json     # Auto-generated KB (created at runtime)
├── chroma_db/                  # Vector store (auto-created)
├── netfix_metrics.db           # SQLite metrics DB (auto-created)
├── .env                        # API keys
├── requirements.txt
└── README.md

---

Setup & Installation

Prerequisites

• Python 3.11
• Groq API key (free at groq.com)

1. Clone the repository

git clone https://github.com/yourusername/NetFixAI.git
cd NetFixAI

2. Create virtual environment

py -3.11 -m venv venv
venv\Scripts\activate        # Windows
source venv/bin/activate     # Mac/Linux

3. Install dependencies

pip install -r requirements.txt

4. Configure environment

Create a .env file in the root directory:

GROQ_API_KEY=your_groq_api_key_here

5. Ingest data

python src/ingest.py

Expected output: Done! All data ingested successfully. Total documents in ChromaDB: 68

6. Run the app

streamlit run app.py

Open http://localhost:8501 in your browser.

---

Sample Queries

What happened to ROUTER-LAB-01 between 08:10 and 08:20?
Why did the BGP session with peer 10.0.0.3 drop?
What is the blast radius of the 5G UPF crash?
Has this type of CPU spike and BGP drop happened before?
What are the fix commands for BGP hold timer expiry on Cisco IOS-XE?
Give me a summary of all open P1 incidents
Which devices in NET-LAB-ALPHA are in WARNING or CRITICAL state?

---

Benchmark Results (10/10 Queries)

All 10 evaluation queries pass with evidence-backed responses:

Query	Result
Root cause of ROUTER-LAB-01 incident	✅ PASS
BGP session drop reason	✅ PASS
WARNING/CRITICAL device list	✅ PASS
SW-LAB-02 blast radius	✅ PASS
Software version mismatch detection	✅ PASS
Historical BGP incident correlation	✅ PASS
Cisco IOS-XE BGP remediation commands	✅ PASS
CRITICAL syslog event extraction	✅ PASS
5G UPF crash blast radius	✅ PASS
Open P1 incident summary	✅ PASS

---

Lab Networks Monitored

Network	Purpose	Devices
NET-LAB-ALPHA	IP routing & feature development	9 devices
NET-LAB-BETA	Regression testing	2 devices
NET-LAB-5G	5G core & slicing validation	4 devices

---

Requirements

langchain
langchain-groq
chromadb
streamlit
pandas
sentence-transformers
networkx
groq
pyvis
python-dotenv
reportlab
numpy

---

Important Notes

• All sample data is synthetic — no real telecom network required
• Never commit your .env file to GitHub
• Always verify AI-generated CLI commands with a senior engineer before execution
• The system reasons from provided data files, not generic LLM knowledge

---

License

For educational and demonstration purposes only.

---

Built with ❤️ for the Student Design & Development Challenge