← Back to Projects
🤖

DataPilot-MCP

GitHub ↗
The Business Problem

Teams have data sitting in SQL databases and CSV files that only technical staff can query, slowing down everyday decisions.

The Solution

A conversational data analytics agent built on the Model Context Protocol (MCP) that answers plain-English questions directly against your database or CSVs.

The Result
  • No query-writing required — ask in plain English, get a chart and narrative insight back
  • MCP used as a production-ready interface layer between LLMs and structured data
  • DuckDB handles in-process analytics for fast, local query execution
Architecture

User Question -> MCP Interface -> LLM -> DuckDB Query -> Chart + Narrative Insight

⚙️ Technical Details For engineers & technical reviewers — click to expand

Ask a question in plain English, get a chart and a narrative insight back — no query-writing required. A conversational data analytics agent built on the Model Context Protocol (MCP) that queries your SQL database or CSV files directly. Built to demonstrate MCP as a production-ready interface layer between LLMs and structured data sources, with DuckDB handling in-process analytics at speed.

MCP DuckDB

Have questions about this project? Ask my AI assistant for details.

Ask AI about this →