Data Dictionary Agent
Link to open source: https://github.com/itsShashankSrivastava/gdg-hackathon-2.0-data-dictionary
Intelligent Data Dictionary Agent
An AI-powered platform that automatically extracts, documents, and analyzes database schemas in minutes—eliminating weeks of manual documentation work.
The Problem:
80% of enterprise databases are poorly documented. Teams waste hours asking DBAs basic questions, new hires take weeks to onboard, and critical knowledge walks out the door when employees leave.
Our Solution:
Connect any database and instantly get:
- Automated Schema Extraction – Complete metadata from PostgreSQL, MySQL, SQL Server, SQLite, Snowflake & Oracle
- Real-Time Data Quality Analysis – Completeness scores, null detection, uniqueness metrics, and freshness tracking
- AI-Generated Business Context – Claude AI transforms technical schemas into plain English descriptions and SQL recommendations
- Conversational Chatbot – Ask "What tables store customer data?" and get instant, context-aware answers
- Multi-Format Export – JSON, Markdown, HTML, and PDF for wikis, Confluence, or stakeholder reports
Business Impact:
- 90% reduction in documentation time
- 75% faster developer onboarding
- 60% fewer DBA interruptions
- ROI in the first year (based on mid-size company savings)
Tech Stack:
React 18 + FastAPI + AWS Bedrock (Claude AI) + Tailwind CSS
This build was uploaded as a hackathon project
