Quickstart

Start querying your databases autonomously

Prerequisites

Before installing Arivu, ensure you have the following configurations available in your environment:

• Python 3.10+ (Arivu depends heavily on the modern AsyncIO stack)
• Operating System: Linux, macOS, or Windows (via WSL2 recommended)
• At least one live database endpoint (PostgreSQL, MySQL, or a local SQLite file)
• An API Key for your preferred Language Model (e.g., OPENAI_API_KEY, ANTHROPIC_API_KEY)

Installation

The Arivu core engine is distributed via PyPI. You can install it natively using pip or your preferred package manager (e.g., poetry, pdm, uv).

pip

pip install arivu

1. Initialize the Arivu Engine

   To begin executing natural language queries against your database, you must first initialize an Arivu engine binding. Ensure that your API keys are visible in your environment .env.

   main.py

   import os
   from arivu import Arivu
   
   # Establish connection with your relational database
   engine = Arivu.connect(
       database_url="postgresql://user:pass@localhost:5432/db",
       llm_provider="openai",
       monitoring=True
   )
   
2. Execute a Pipeline Request

   Once the engine proxy is created, you can execute dynamic text-to-SQL logic by sending a natural language prompt.

   main.py

   # Formulate a natural language query
   q = engine.query("What were our top 3 highest revenue products last quarter?")
   
   # Run the intelligent execution pipeline
   response = engine.run_pipeline(q)
   
   print(response.get("result"))
   

Launching the Web Dashboard

If you prefer a visual observability interface, Arivu provides an underlying Next.js dashboard template that binds tightly to your core configuration.

Wait for our comprehensive Dashboard Guide on how to spin up the local graphical user interface!