SQL Query Prompt: Write Complex Queries

This sql query prompt writes optimized queries from plain-English requests with explanations and indexes. Describe the tables, paste into any AI tool.

The prompt

Role:
Act as a database engineer with 10 years of experience in query optimization.

Context:
Database: [DATABASE_TYPE]. Table schemas: [TABLE_SCHEMAS]. Request in plain English: [REQUEST]. Performance needs: [PERFORMANCE_NEEDS] (e.g., runs on 10M rows, must finish under 2s).

Task:
1. Write the SQL query that fulfills [REQUEST].
2. Explain the query logic step by step: joins, filters, aggregations.
3. Suggest indexes that speed it up, with CREATE INDEX statements.
4. Note edge cases: NULLs, duplicates, timezone handling.
5. Provide a simpler alternative if the request allows a trade-off.
6. Estimate the query plan shape (scan vs index seek) for the largest table.

Format:
- SQL (formatted code block)
- Logic explained
- Recommended indexes (code block)
- Edge cases
- Simpler alternative

Constraints:
- Use standard SQL compatible with [DATABASE_TYPE]; note dialect-specific functions.
- Never use SELECT * in the final query.
- Parameterize values; no hardcoded literals except in examples.
- The query must be read-safe: no writes, no schema changes.

Variables to fill in

Replace each [variable] in the prompt with your own details:

[DATABASE_TYPE] [TABLE_SCHEMAS] [REQUEST] [PERFORMANCE_NEEDS]

Example input

DATABASE_TYPE: PostgreSQL 16; TABLE_SCHEMAS: users(id, created_at), orders(id, user_id, total, status, created_at); REQUEST: monthly revenue per new vs returning customer for 2025; PERFORMANCE_NEEDS: 8M orders, under 3s

How to use this prompt

Copy the prompt. Paste table schemas and describe the request plainly. Run in any AI tool, then test on a staging database first.

Tips for the best results

Tip: Always EXPLAIN the generated query before running it on large tables.
Tip: Include row counts per table — the AI picks better join strategies with that context.
Tip: Ask for the query with and without window functions to compare readability.