Simple Data Dictionary (AI-Powered)

This enhanced data dictionary template helps you quickly generate field-level metadata for your data models or marts using AI-driven logic. Just add your column name and associated entity, and the template infers both DATA_TYPE and a business-friendly DESCRIPTION. Whether you’re prototyping a mart or standardizing an existing model, this tool removes the initial documentation drag so you can stay in the flow.

How It Works

1. Add your FIELD_NAME
2. Specify the ENTITY it belongs to (e.g., Customer, Order, Event)
3. Let AI infer the DATA_TYPE and generate a context-rich DESCRIPTION
4. Review, tweak, and evolve — AI gives you a solid first draft

⚠️ AI isn’t flawless. Expect to fine-tune as needed. You can also adjust the embedded prompt logic to accommodate new naming conventions or domain-specific patterns.

Columns Included

1. ID → Unique identifier for tracking rows
2. FIELD_NAME → The name of the field/column in your data model
3. DATA_TYPE → AI-generated guess of the appropriate data type
4. DESCRIPTION → AI-generated field description based on name + entity

🎯 Best For

• Data engineers and analysts kickstarting a new data mart
• Consultants building fast client documentation for handoff
• Anyone standardizing a Kimball-style dimensional model
• Teams wanting an AI copilot for documentation — but still in control

💡 Pro Tips

• Use meaningful field names and entities to improve inference quality
• Incorporate as a first pass in your dbt docs or internal Confluence
• Keep a list of oddball fields that consistently confuse the AI — and refine the prompt accordingly