Introduction

Welcome to the Arivu Documentation

What is Arivu?

Arivu is an open-source framework designed for building autonomous database agents. It bridges the gap between raw data warehouses and large language models (LLMs) by offering a dual-layer approach:

1. A Powerful Python SDK: Allows engineers to orchestrate text-to-SQL pipelines, hook up custom agents, and seamlessly inject memory layers.
2. A Sleek Dashboard UI: A Next.js-based no-code observability platform designed for teams to instantly connect dialects, select cloud models, execute conversational prompts, and inspect pipeline execution latency.

Explore

Dashboard

Explore your data with a beautifully crafted, theme-aware user interface.

Python SDK

Embed Arivu directly into your own applications using our Python SDK.

MCP Server

Connect Arivu directly to Claude Desktop using the Model Context Protocol.

Memory System

Understand how Arivu persists sessions, traces, and feedback signals.

Highlights

Generative Visualizations

Arivu automatically detects tabular data and generates interactive, theme-aware charts using Thesys C1.

Self-Healing SQL

If the LLM generates invalid SQL, the query_verifier node automatically loops back and corrects the syntax before execution.

Admin Gatekeeper

Destructive operations (like DROP or DELETE) are flagged by the pipeline and safely paused until an admin explicitly approves them.

Semantic Caching

Repeated queries are instantly served from a TTL cache, bypassing the LLM layer completely for lightning-fast responses.