AI Data Dictionary
Input: Raw Schema
CREATE TABLE customers ( id INT PRIMARY KEY, email VARCHAR(255), created_at TIMESTAMP );
Output: Enriched Doc
id (Integer)
Unique identifier for customer records.
Confidence: 99%
email (String)
Customer's primary contact email.
Confidence: 95%PII: Email
Executive Summary Generation
For non-technical stakeholders, we generate a high-level overview of the database's business function.
Database Overview: E-Commerce Production
This database supports the core e-commerce platform, handling Order Processing, Inventory Management, and Customer Profiles.
47
Tables
2.3M
Customers
5
Business Domains
Key Insight: The schema is well-normalized (3NF) but the orders table is growing rapidly (8.7M rows) and may require partitioning soon.
Multi-Format Exports
Interactive HTML
Searchable, responsive website (like this one!) for your internal wiki. Shareable via secure link.
Excel / CSV
Structured spreadsheets for business analysts and bulk manual editing.
JSON / Markdown
Machine-readable formats for developers to commit to Git repos.
PDF Reports
Polished, print-ready documents for compliance audits (SOC 2, ISO).
Confidence Scoring
| Score Range | Meaning | Recommended Action |
|---|---|---|
| 90-100% | High Confidence. Clear naming & data patterns. | Light skim. Usually accurate. |
| 70-89% | Medium Confidence. Some ambiguity. | Quick verification recommended. |
| < 70% | Low Confidence. Obscure naming (e.g., col_1). | Requires human review. |