Getting Started

import cognee
import asyncio

async def main():
    # Add some text data
    text = """
    Artificial Intelligence is transforming healthcare through machine learning 
    algorithms that analyze medical images and predict patient outcomes. 
    Deep learning models excel at pattern recognition in radiology.
    """
    
    result = await cognee.add(text)
    print("Data added:", result)

asyncio.run(main())

import cognee
import asyncio

async def main():
    # Process all added data
    result = await cognee.cognify()
    print("Knowledge graph created:", result)

asyncio.run(main())

import cognee
import asyncio
from cognee import SearchType

async def main():
    # Search with natural language
    results = await cognee.search(
        query_text="What is artificial intelligence?",
        query_type=SearchType.GRAPH_COMPLETION
    )
    
    print(results[0])

asyncio.run(main())

import cognee
import asyncio

async def main():
    # Add new information
    new_info = "AI also helps in drug discovery and personalized medicine."
    result = await cognee.add(new_info)
    print("New data added:", result)
    
    # Incremental processing (only new data)
    cognify_result = await cognee.cognify()
    print("Processing result:", cognify_result)
    
    print("Memory updated successfully!")

asyncio.run(main())

import cognee
import asyncio

async def main():
    # Delete specific data by ID
    delete_result = await cognee.delete(data_id="abc123-def456-ghi789")
    print("Delete result:", delete_result)
    
    # Or clear all memory
    # prune_result = await cognee.prune()
    # print("Prune result:", prune_result)
    
    print("Data deleted successfully!")

asyncio.run(main())

import cognee
import asyncio
import os
from cognee import SearchType

async def main():
    # 1. Setup (add your API key)
    os.environ["LLM_API_KEY"] = "your-openai-api-key"
    
    # 2. Add knowledge
    knowledge_base = """
    Cognee is an AI memory layer that transforms unstructured data into 
    knowledge graphs. It uses large language models to extract entities 
    and relationships, creating searchable semantic networks.
    
    Key features include:
    - Automatic entity extraction
    - Relationship mapping  
    - Multi-modal data support
    - Natural language querying
    - Incremental learning
    """
    
    print("📚 Adding knowledge...")
    await cognee.add(knowledge_base)
    
    # 3. Process into knowledge graph
    print("🧠 Building knowledge graph...")
    await cognee.cognify()
    
    # 4. Query the knowledge
    print("🔍 Searching knowledge...")
    
    # Natural language question
    results = await cognee.search(
        query_text="What are the key features of Cognee?",
        query_type=SearchType.GRAPH_COMPLETION
    )
    
    print("\n✨ Answer:")
    print(results[0])
    
    # Get related concepts
    insights = await cognee.search(
        query_text="Cognee features",
        query_type=SearchType.INSIGHTS
    )
    
    print("\n🔗 Related Concepts:")
    for insight in insights[:3]:
        print(f"- {insight}")

if __name__ == "__main__":
    asyncio.run(main())

import cognee
import asyncio

async def main():
    # Add memories with context
    result1 = await cognee.add(
        "User prefers technical documentation",
        user_id="alice",
        node_set="onboarding-001"
    )
    print("Added memory 1:", result1)
    
    result2 = await cognee.add(
        "User is interested in machine learning",
        user_id="alice", 
        node_set="onboarding-001"
    )
    print("Added memory 2:", result2)
    
    # Process with context
    cognify_result = await cognee.cognify()
    print("Processing result:", cognify_result)
    
    # Search with different scopes
    user_memories = await cognee.search(
        query_text="What do you know about the user?",
        user_id="alice"
    )
    
    node_set_memories = await cognee.search(
        query_text="What happened in this session?",
        user_id="alice",
        node_set="onboarding-001"
    )
    
    print("User memories:", len(user_memories))
    print("Node Set memories:", len(node_set_memories))
    print("Sample user memory:", user_memories[0][:100] + "...")

asyncio.run(main())

import cognee
import os
import asyncio

async def main():
    # Fully local configuration
    os.environ["LLM_PROVIDER"] = "ollama"
    os.environ["LLM_MODEL"] = "llama3.2"
    os.environ["EMBEDDING_PROVIDER"] = "fastembed"
    os.environ["EMBEDDING_MODEL"] = "sentence-transformers/all-MiniLM-L6-v2"
    
    print("Setting up local processing...")
    
    # Your code here
    add_result = await cognee.add("Local knowledge processing with Cognee")
    print("Data added:", add_result)
    
    cognify_result = await cognee.cognify()
    print("Processing complete:", cognify_result)
    
    results = await cognee.search("Tell me about local processing")
    print("Search result:", results[0])

asyncio.run(main())