Data Dictionary Generator — Document & Govern
A live, gated demo: upload a CSV and auto-generate a governance-ready data dictionary — inferred business meaning per column, PII flags, and sensitivity ratings — downloadable as Markdown. LLM-agnostic; this instance uses Gemini.
Overview
Data Dictionary Generator produces governance-ready documentation from a raw dataset. Each column is profiled (type, null rate, distinct count, example values), then an AI governance reviewer infers its business meaning, flags personally identifiable information (PII), and assigns a sensitivity level (none / low / medium / high). The result is a structured data dictionary, downloadable as Markdown for a data catalog or wiki. LLM-agnostic — runs on a local model (Ollama / Llama) or any cloud LLM; this deployment uses Gemini. Data is processed in-memory and expires automatically.
Problem Solved
Most datasets arrive undocumented, and writing a data dictionary by hand is tedious and often skipped — leaving teams unsure what columns mean or which hold sensitive data. This demo bootstraps that documentation in seconds, including PII and sensitivity flags that are essential for governance and compliance.
Capabilities
- Per-column profiling with example values
- AI-inferred business meaning and descriptions
- PII detection and sensitivity rating (none/low/medium/high)
- Downloadable Markdown data dictionary
- CSV upload or built-in sample datasets
- In-memory processing, rate limiting, gated access