/sql_study

Practice intermediate and advanced SQL exercises using my curated ChatGPT prompt.

Primary LanguageJupyter Notebook

SQL Interview Exercise Generator Prompt

This repository contains a ready-to-use prompt for ChatGPT that generates intermediate to advanced SQL exercises tailored for data analysts, BI developers, data integration specialists, analytics engineers, and SQL specialists.

🎯 Purpose

The prompt is designed to help users practice SQL for interviews, skill sharpening, or exam preparation. Each execution generates a unique, realistic exercise with:

  • Context-specific schema and tables, using short randomized schema names (contextXYZ).
  • Realistic sample data, including duplicates, NULLs, and tricky edge cases.
  • Exercises requiring 1–2 SQL techniques only (CTEs, Window Functions, Aggregates, Subqueries, JOINs).
  • Narrative problem statements written as if requested by a client.
  • Expected Results displayed as mock tables with representative rows and column types.

🛠 How to Use

  1. Open a new ChatGPT session.

  2. Copy and paste the full prompt into the session.

  3. ChatGPT will generate a complete SQL exercise, including:

    • DDL: Table creation with primary keys, foreign keys, and idempotent DROP/DELETE commands.
    • DML: Sample data populated with realistic logic traps.
    • Problem Statement: Written as a narrative for practical context.
    • Expected Results: Mock table layout reflecting the final output.

  4. Run the SQL code in any SQL Server-compatible environment, such as:

    • SQL Server (on-premises)
    • SQL Server Management Studio (SSMS)
    • Visual Studio Code with Jupyter Notebook integration

The exercises are idempotent, so running the code multiple times will not produce duplicates or errors.

🔄 Variability

Each execution of the prompt generates a different exercise, varying:

  • Number of tables (1–2)
  • SQL techniques required (maximum 2 per exercise)
  • Table structures and Expected Results layouts
  • Sample data and business scenarios, including edge cases

This ensures daily practice with unique, realistic challenges.

âš  Notes

  • The prompt is designed for intermediate to advanced SQL practice.
  • Each exercise is solvable within 15–30 minutes.
  • Only 1–2 tasks/questions per exercise are included, focusing on a single SQL problem or clear combination of techniques.
  • Schema, table, and column names are randomized to avoid collisions with previous exercises.

The prompt:

see https://github.com/pedrocarvalho01/sql_study/blob/main/chatgpt_prompt.txt