Skip to content

LouieAI Python Client

Welcome to the LouieAI Python client library documentation.

LouieAI is Graphistry's genAI-native investigation platform. This library allows Python applications to interact with LouieAI via its API, leveraging Graphistry authentication.

Video: Louie <> Graphistry - Python edition!

Install & Go

pip install louieai

import graphistry
import louieai

# Configure servers and authenticate
graphistry.register(
    api=3, 
    server="hub.graphistry.com",  # Your Graphistry server
    username="alice@example.com", 
    password="<your_password>"
)

# Create Louie interface with server configuration
lui = louieai(server_url="https://den.louie.ai")  # Your Louie server

# Start analyzing
lui("Show me patterns in customer behavior")
print(lui.text)  # Natural language insights
df = lui.df      # Data as pandas DataFrame

Powerful Options

# Choose your authentication method
import louieai

# Option 1: Environment variables (great for notebooks)
# export GRAPHISTRY_USERNAME="your_username" 
# export GRAPHISTRY_PASSWORD="your_password"
lui = louieai()  # Auto-detects credentials

# Option 2: Specific servers and organizations
g = graphistry.register(
    api=3,
    server="hub.graphistry.com",  # or "your-company.graphistry.com"
    username="your_username",
    password="your_password",
    org_name="your-org"  # Optional: specify organization
)
lui = louieai(g, server_url="https://den.louie.ai")  # or your enterprise URL

# Control data visibility
lui = louieai(share_mode="Organization")  # Share within your org
lui("Analyze sales trends", share_mode="Private")  # Override per query

# Enable AI reasoning traces
lui.traces = True
lui("Complex analysis requiring step-by-step reasoning")

# Access conversation history
previous_result = lui[-1]  # Last response
older_df = lui[-2].df      # DataFrame from 2 queries ago

Need more options? See our guides:

Key Features

  • Notebook-friendly API: Streamlined lui() interface for Jupyter notebooks
  • Thread-based conversations: Maintain context across multiple queries
  • Multiple response types: Handle text, DataFrames, visualizations, and more
  • 40+ Specialized Agents: Choose from database-specific, visualization, and analysis agents
  • Real-time streaming: See responses as they're generated in Jupyter notebooks
  • Natural language interface: Access all Louie capabilities through simple prompts
  • Auto-refresh authentication: Automatically handles JWT token expiration
  • Multiple auth methods: Works with existing Graphistry sessions or direct credentials

Available Agents with Semantic Understanding

LouieAI provides specialized agents that learn and understand your data:

  • General Purpose: LouieAgent (default), TextAgent, CodeAgent
  • Databases with Semantic Layer: DatabricksAgent, PostgresAgent, MySQLAgent, SnowflakeAgent, BigQueryAgent
  • Agents learn your schema, relationships, and business context
  • Generate complex queries from natural language using semantic understanding
  • Search & Analytics: SplunkAgent, OpenSearchAgent, KustoAgent
  • Visualizations: GraphAgent, PerspectiveAgent, KeplerAgent, MermaidAgent
  • Direct Execution: PassthroughAgent variants for each database (no AI interpretation)
# Use the default conversational agent
lui("Analyze security incidents from last week")

# Database agent with semantic understanding
lui("Show me customer churn trends", agent="DatabricksAgent")
# The agent understands your schema and business definitions of "churn"

# Natural language leveraging learned semantics
lui("Which products have anomalous return rates?", agent="PostgresAgent") 
# Agent knows your product hierarchy, return policies, and what's "anomalous"

# Direct SQL when you need exact control
lui("SELECT * FROM auth_logs WHERE status='failed'", agent="PostgresPassthroughAgent")

See the complete Agents Reference for all available agents and usage examples.

Getting Started

New to LouieAI? Start here:

  1. Installation - Install the LouieAI Python client
  2. Authentication - Set up authentication with PyGraphistry
  3. Quick Start - Make your first queries and explore features

User Guides

Ready to dive deeper? These guides cover common use cases and advanced features:

API Reference

Complete technical documentation:

Developer Resources

Contributing to LouieAI or setting up for development:

Support