Lightning.AI

​

Overview

• Graph-Enhanced RAG - Leverage FalkorDB’s graph database capabilities to enhance your RAG applications with contextual relationships
• Managed Infrastructure - Lightning.AI handles the infrastructure, so you can focus on building your application
• Easy Deployment - Get started quickly with pre-configured environments
• Scalable - Scale your applications as your needs grow

​

Getting Started with FalkorDB on Lightning.AI

​

Access the Environment

1. Visit the FalkorDB Lightning.AI Environment
2. Sign in to your Lightning.AI account or create one if needed
3. Fork or use the environment to start building your application

​

Environment Features

• Pre-configured FalkorDB Instance - Ready-to-use graph database
• Sample Code and Notebooks - Examples demonstrating graph-enhanced RAG patterns
• Required Dependencies - All necessary libraries and tools pre-installed
• Interactive Development - Jupyter notebooks for interactive exploration

​

Use Cases

​

Advanced RAG with Graph Context

from falkordb import FalkorDB

# Connect to FalkorDB
db = FalkorDB(host='localhost', port=6379)

# Select a graph for your knowledge base
graph = db.select_graph('knowledge_base')

# Create entities and relationships
graph.query("""
    CREATE (d:Document {id: 'doc1', content: 'FalkorDB is a graph database'}),
           (t:Topic {name: 'Graph Databases'}),
           (d)-[:RELATES_TO]->(t)
""")

# Query with graph context for RAG
result = graph.query("""
    MATCH (d:Document)-[:RELATES_TO]->(t:Topic {name: $topic})
    RETURN d.content
""", {'topic': 'Graph Databases'})

​

Building GenAI Applications

1. Knowledge Graph Construction - Build structured knowledge from unstructured data
2. Context-Aware Retrieval - Use graph relationships to find relevant information
3. Enhanced Generation - Provide LLMs with rich, connected context
4. Citation and Traceability - Track information sources through graph relationships

​

Integration Patterns

​

Pattern 1: Graph-Enhanced Document Retrieval

# Store documents with metadata and relationships
graph.query("""
    CREATE (d:Document {id: $doc_id, content: $content, embedding: $embedding}),
           (a:Author {name: $author}),
           (t:Topic {name: $topic}),
           (d)-[:WRITTEN_BY]->(a),
           (d)-[:ABOUT]->(t)
""", params={'doc_id': 'doc1', 'content': '...', 'embedding': [0.1, 0.2, 0.3],  # Example embedding vector
             'author': 'John Doe', 'topic': 'AI'})

# Retrieve with graph context
result = graph.query("""
    MATCH (d:Document)-[:ABOUT]->(t:Topic {name: $topic})
    MATCH (d)-[:WRITTEN_BY]->(a:Author)
    RETURN d.content, a.name, t.name
""", {'topic': 'AI'})

​

Pattern 2: Entity Relationship Extraction

# Extract entities and relationships from text
graph.query("""
    MERGE (e1:Entity {name: $entity1, type: $type1})
    MERGE (e2:Entity {name: $entity2, type: $type2})
    MERGE (e1)-[r:RELATES_TO {relation: $relation}]->(e2)
""", params={'entity1': 'FalkorDB', 'type1': 'Database',
             'entity2': 'Graph', 'type2': 'Concept',
             'relation': 'is_type_of'})

# Query relationships for context
result = graph.query("""
    MATCH (e1:Entity {name: $entity})-[r]->(e2:Entity)
    RETURN e2.name, type(r), e2.type
""", {'entity': 'FalkorDB'})

​

Resources

​

Documentation and Guides

• GraphRAG Blog Post

​

Next Steps

1. Explore the Environment - Try the FalkorDB Lightning.AI environment
2. Build Your First Graph - Create a simple knowledge graph
3. Integrate with LLMs - Connect FalkorDB with your favorite LLM API
4. Deploy Your Application - Use Lightning.AI’s deployment features to share your work

​

Additional Deployment Options

• FalkorDB Cloud - Managed cloud service
• Railway Deployment - Quick deployment on Railway
• Kubernetes - Production-grade orchestration
• Docker - Local or self-hosted deployment

​

Frequently Asked Questions