A demonstration of how natural language interfaces can reduce the distance between business questions and data answers — without requiring analyst intervention for every query.
Back to Projects
T
In DevelopmentTool2026
Text-to-SQL-to-Visualisation Engine
Natural language querying for enterprise data — ask a question, get a chart.
The Problem
Business users have questions. Analysts have answers. But the queue between them is long, the turnaround is slow, and most ad-hoc questions don't justify the analyst's time. The alternative — training non-technical users to write SQL — doesn't scale. The result is a data team that spends most of its time answering simple questions instead of solving hard ones.
The Solution
A conversational interface that accepts a plain-English question, generates the correct SQL using an LLM with schema context, executes it safely, and returns a chart alongside the underlying data. The system handles schema disambiguation, query validation, and chart type selection automatically — no SQL knowledge required from the end user.
Tech Stack
PythonLangChainOpenAI APISQLPlotly