Knowledge Graph Use Cases

Technical workflows demonstrating NetIntel-OCR v0.1.17's Knowledge Graph capabilities through real-world scenarios.

Available Use Cases

📊 Network Infrastructure Migration

Transform complex network documentation into actionable migration plans with dependency analysis and risk assessment.

Key Capabilities: - Dependency mapping across 3,847 components - Hidden connection discovery - Migration wave planning - Cloud compatibility validation - Risk assessment and validation

Metrics: - Analysis time: 3 months → 2 days - Accuracy: 94% vs 67% manual - Hidden dependencies found: 23

---

🔒 Security Compliance Audit

Automate multi-tenant security compliance for Managed Security Service Providers with configuration drift detection and automated remediation.

Key Capabilities: - Multi-framework compliance checking - Configuration drift analysis - Access path tracing - Automated remediation scripts - Zero-trust validation

Metrics: - Audit time: 3 weeks → 2 days - Compliance score: 78% → 92% - Violations detected: 96% accuracy

---

🚨 Intelligent Incident Response

Real-time incident correlation with automated runbook execution and root cause analysis.

Key Capabilities: - Multi-source alert correlation - Automated runbook matching - Root cause analysis - Pattern recognition - Prevention recommendations

Metrics: - MTTR: 45 min → 15 min (67% reduction) - Auto-resolution: 78.7% of incidents - Revenue protected: $3.2M annually

---

💬 Customer Service Chat Intelligence

Enable intelligent customer service for complex enterprise telecom offerings with real-time pricing and availability from ServiceNow and Salesforce.

Key Capabilities: - Multi-source data integration (ServiceNow CMDB, Salesforce CRM) - Service matching and recommendation - Dynamic pricing and bundling - Location-based availability checking - Automated chat responses with 94% accuracy

Metrics: - Response time: 1.8 seconds average - Lead qualification: +45% improvement - Agent productivity: +210% - Quote accuracy: 99.2%

---

Performance Comparison

Query Performance Across Use Cases

Operation	Traditional	Vector Only	Graph Only	KG Embeddings	Hybrid KG
Dependency Analysis	2.3s	450ms	120ms	85ms	180ms
Compliance Check	5.1s	780ms	230ms	140ms	290ms
Incident Correlation	8.7s	920ms	340ms	190ms	410ms
Accuracy	62%	71%	83%	87%	94%

Storage Efficiency

Traditional Approach:
  Separate databases: 15.3 GB
  Redundant data: 4.7 GB
  Total: 20 GB

NetIntel-OCR KG:
  FalkorDB: 3.2 GB
  Milvus: 8.1 GB
  Total: 11.3 GB
  Savings: 43.5%

Technical Architecture

Knowledge Graph Stack

┌─────────────────────────────────────┐
│         PDF Documents               │
│  (Network, Security, Operations)    │
└────────────┬────────────────────────┘
             │
┌────────────▼────────────────────────┐
│      NetIntel-OCR v0.1.17          │
│  • Diagram Detection                │
│  • Table Extraction                 │
│  • Entity Recognition               │
└────────────┬────────────────────────┘
             │
    ┌────────┴────────┬──────────┐
    ▼                 ▼          ▼
┌─────────┐    ┌──────────┐  ┌─────────┐
│FalkorDB │    │  PyKEEN  │  │ Milvus  │
│ Graph   │    │Embeddings│  │ Vectors │
└─────────┘    └──────────┘  └─────────┘
    │                 │          │
    └────────┬────────┴──────────┘
             │
┌────────────▼────────────────────────┐
│       Hybrid Retrieval              │
│  • Query Classification             │
│  • Strategy Selection               │
│  • Result Fusion                    │
└─────────────────────────────────────┘

Getting Started

Prerequisites

# Install NetIntel-OCR
pip install netintel-ocr==0.1.17

# Configure external Ollama server
export OLLAMA_HOST="http://your-ollama-server:11434"

# Verify Ollama has required models
curl $OLLAMA_HOST/api/tags | jq '.models[].name'
# Required: qwen2.5vl:7b, gemma3:4b-it-qat, qwen3-embedding:8b

# Start required graph/vector services
docker-compose up -d falkordb milvus

# Verify installation
netintel-ocr kg check-requirements

Quick Commands

# Process documents with KG
netintel-ocr process batch --kg-model RotatE documents/*.pdf

# Query knowledge graph
netintel-ocr kg query --entity "System-Name" --max-depth 3

# Run compliance check
netintel-ocr kg compliance --framework PCI-DSS-v4.0

# Correlate incident
netintel-ocr kg correlate --incident alert.json

Choosing the Right Use Case

When to Use Each Approach

Scenario	Recommended Use Case	Key Benefits
Cloud migration planning	Migration	Dependency mapping, risk assessment
Compliance audit	Compliance	Automated validation, remediation
Service outage	Incident Response	Fast correlation, runbook automation
Network redesign	Migration	Impact analysis, validation
Security assessment	Compliance	Vulnerability detection, drift analysis
Performance issues	Incident Response	Pattern recognition, RCA

Advanced Features

Knowledge Graph Capabilities

• Entity Extraction: Automatic identification of network components, services, and relationships
• Embedding Training: 8 PyKEEN models for different relationship types
• Hybrid Retrieval: Combines graph traversal, vector similarity, and KG embeddings
• Query Routing: Automatic selection of optimal retrieval strategy
• Real-time Updates: Continuous learning from new documents

Integration Options

• REST API: HTTP endpoints for all operations
• Python SDK: Native Python library for custom workflows
• CLI Tools: Command-line interface for automation
• Docker/K8s: Container deployment for scalability

Support and Resources

• Documentation: Knowledge Graph Guide
• Quick Start: Getting Started
• API Reference: API Documentation
• Troubleshooting: Support Guide

---

Select a use case above to explore detailed technical workflows and command examples.