databricks-solutions

Dbx Unifiedchat

Name: Dbx Unifiedchat
Author: databricks-solutions

Communication community

SOURCE https://github.com/databricks-solutions/dbx-unifiedchat

Description

dbx-unifiedchat is a custom agent stack with an app interface for NL2SQL cross-domain chat on Databricks as a Databricks Solution.

Installation

This entry records only its repository, not the path inside it, so there is no exact command to give. Open the source below and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

README

![DBX-UnifiedChat Logo](docs/logos/dbx-unifiedchat-logo-pacman-eating-data.png)

DBX-UnifiedChat - Databricks Unified Chat

A multi-agent system for intelligent cross-domain data queries built with LangGraph, Databricks Genie, Lakebase, and Claude models/skills on Databricks Platform.

[![Python 3.11+](https://img.shields.io/badge/python-3.11+-blue.svg)](https://www.python.org/downloads/) [![License](https://img.shields.io/badge/license-Databricks-blue.svg)](LICENSE.md) [![LangGraph](https://img.shields.io/badge/LangGraph-0.0.30-green.svg)](https://github.com/langchain-ai/langgraph) [![LangChain](https://img.shields.io/badge/LangChain-≥0.1.0-green.svg)](https://github.com/langchain-ai/langchain) [![MLflow](https://img.shields.io/badge/MLflow-≥3.6.0-orange.svg)](https://mlflow.org/) [![Databricks SDK](https://img.shields.io/badge/Databricks%20SDK-≥0.20.0-red.svg)](https://github.com/databricks/databricks-sdk-py) [![Pydantic](https://img.shields.io/badge/Pydantic-≥2.0.0-blue.svg)](https://github.com/pydantic/pydantic) [![Claude Models](https://img.shields.io/badge/Claude-Sonnet%2FHaiku_4.5-purple.svg)](https://www.anthropic.com/claude) [![Claude Skills](https://img.shields.io/badge/Claude-Skills-purple.svg)](https://www.anthropic.com/claude)

Overview

Organizations struggle to query data across multiple domains and data sources, requiring deep SQL expertise and knowledge of complex data schemas. **Databricks Unified Chat** solves this by providing an intelligent multi-agent system that routes natural language queries to the appropriate data sources, synthesizes results, and delivers comprehensive answers.

Built on LangGraph, Databricks Genie and Lakebase, this solution enables business users to ask questions spanning multiple data domains without needing to understand the underlying data architecture or write complex SQL queries.

The supported application workflow now lives under `agent_app/`, which contains the Databricks App bundle, backend agent runt