FrankChen021/datastoria

What it solves

DataStoria provides an AI-enhanced interface for managing ClickHouse clusters. It simplifies the process of writing complex SQL queries, diagnosing performance bottlenecks, and creating data visualizations, reducing the reliance on manual SQL expertise and deep internal knowledge of ClickHouse system logs.

How it works

The tool acts as a console that connects directly to your ClickHouse server. It uses LLMs (via Vercel AI SDK) to translate natural language requests into optimized SQL and visualization specifications. To maintain privacy, it executes queries locally from the browser to the server, ensuring that actual data results never leave the user's machine.

Who it’s for

It is designed for data engineers, database administrators, and analysts who work with ClickHouse and want to accelerate their query generation, cluster monitoring, and performance optimization workflows.

Highlights

• AI-Powered Querying: Converts plain English into optimized ClickHouse SQL and generates chart specifications for time series, bar, and pie views.
• Cluster Diagnostics: Includes a system log inspector with timeline views and topology graphs to analyze query execution.
• Multi-Cluster Management: Supports managing multiple clusters and nodes from a single dashboard with real-time metrics.
• Privacy-First Architecture: Operates with 100% local execution of SQL queries and allows users to bring their own LLM API keys.
• Agentic Capabilities: Supports a skill-based agent architecture to maintain context and save tokens across multi-turn conversations.