Generate an optimized, production-ready database schema for any app or SaaS project — with tables, relationships, indexes, and foreign keys explained clearly.
📋 The Prompt
Act as a senior database architect.
I am building: [e.g. a project management SaaS / e-commerce platform /
social media app]
Tech stack: [e.g. PostgreSQL / MySQL / MongoDB]
I need a complete database schema with the following entities:
[e.g. Users, Workspaces, Projects, Tasks, Comments, Subscriptions]
For each table provide:
- All columns with data types
- Primary keys and foreign keys
- Indexes for fast querying (explain WHY each index is needed)
- Timestamp fields (created_at, updated_at)
- Any ENUM types needed
- Constraints and validations at DB level
Also include:
- Raw SQL CREATE TABLE statements I can run directly
- An entity relationship diagram description (text format)
- Top 3 performance considerations for this schema at scale
- Any columns I might be forgetting for a production app
Format output as clean SQL code blocks with comments.
🚀 How to Use This Prompt
Copy the prompt above and paste it into your preferred AI tool — Midjourney, DALL-E, ChatGPT, or Stable Diffusion. Replace all [bracketed placeholders] with your specific values for best results.
💡 Tips for Best Results
- Be as specific as possible when filling in the placeholders.
- Experiment with different values for varied, unique results.
- For Midjourney, add
--v 6 --ar 16:9at the end. - Run the same prompt multiple times — AI generates unique outputs each run.
More Coding Prompts
Coding
REST API Builder with Authentication : Build a Complete REST API with Auth — AI Prompt
Generate a production-ready REST API with JWT authentication, input validation, and error ...
Coding
Brutally Honest AI Advisor Prompt for Clarity, Growth & Strategic Thinking
Use this brutally honest AI advisor prompt to challenge weak thinking, expose blind spots,...
Coding
AI Code Refactoring Prompt — Clean Up Messy Code Instantly
Transform messy, unreadable, or inefficient code into clean, maintainable, production-read...