Skip to content

Instantly share code, notes, and snippets.

@ned1313
Created February 4, 2026 12:01
Show Gist options
  • Select an option

  • Save ned1313/e40dcbec354295e86b9c56a320ef2b59 to your computer and use it in GitHub Desktop.

Select an option

Save ned1313/e40dcbec354295e86b9c56a320ef2b59 to your computer and use it in GitHub Desktop.
AI-900 Question Creation Prompt

AI-900 Azure AI Fundamentals Question Generation

You are an expert Azure AI professional and certification trainer creating high-quality AI-900 Microsoft Certified: Azure AI Fundamentals practice questions. You will be generating 50 questions for the following domain specific areas.

Domain-Specific Focus Areas:

  • Describe Artificial Intelligence workloads and considerations - 20%
  • Describe fundamental principles of machine learning on Azure - 20%
  • Describe features of computer vision workloads on Azure - 20%
  • Describe features of Natural Language Processing (NLP) workloads on Azure - 20%
  • Describe features of generative AI workloads on Azure - 20%

Use the percentage of each area to create a representative mix of questions.

Requirements:

  • Target the AI-900 Microsoft Certified: Azure AI Fundamentals certification level
  • Create a mix of scenario-based questions that test practical application, and more straightforward questions that test knowledge
  • Include real-world situations a AI Engineer would encounter
  • Ensure questions test deep understanding of Azure AI concepts and best practices
  • Save the file using the naming convention: Azure AI-YYYYMMDD-Version##.json (e.g. Azure AI-20260201-Version01.json)
  • Save the generated JSON file to the certification folder for AI-900

Question Format (JSON):

{
  "question": "Detailed scenario-based question (2-4 sentences)",
  "correct_answer": "Precise, actionable correct answer",
  "wrong_answers": ["Plausible but incorrect option", "Another realistic distractor", "Common misconception", "Related but wrong approach"],
  "explanation": "Comprehensive explanation (3-5 sentences) covering why the correct answer is right and why others are wrong. Include key concepts and best practices. Also include single sentence explanations of why the distractors were incorrect.",
  "references": ["https://learn.microsoft.com/en-us/credentials/certifications/resources/study-guides/ai-900 (official documentation URL)", "https://learn.microsoft.com/ (official learning platform URL)"],
  "category": "Describe Artificial Intelligence workloads and considerations"
}

Content Guidelines:

  • Use current Azure AI services and features (as of 2025)
  • Reference official Microsoft documentation and learning resources from the URLs provided
  • Include specific Azure AI resource names, settings, and configurations
  • Test understanding of service limits, pricing tiers, and architectural decisions
  • Cover troubleshooting scenarios and performance optimization
  • Include security, compliance, and governance considerations
  • Reference real platform workflows and CLI/API commands
  • Use MCP servers or equivalent terminology where applicable to access resources

Quality Standards:

  • Questions should require 1-2 levels of Azure AI knowledge to answer correctly
  • Correct answers should vary in length and word choice
  • Distractors should be realistic options an inexperienced professional might choose
  • Distractors and correct answers should be of similar length
  • Explanations must reference official Microsoft documentation
  • Each question should teach something valuable beyond just testing knowledge
  • Avoid questions with obvious answers or trick questions

IMPORTANT - Post-Generation Validation:

After generating the JSON, it's critical to validate the output before use:

  1. Save the generated JSON to a file with a descriptive name
  2. Run validation using: .\Validate-QuizJSON.ps1 -FilePath "your-file.json"
  3. Fix any errors or warnings reported by the validation script
  4. Re-validate until the script reports "VALIDATION PASSED"

The validation script checks for:

  • JSON syntax correctness
  • Required field presence and proper data types
  • Answer uniqueness and quality
  • Content structure and formatting
  • Common issues like field naming errors

Generate questions that an experienced AI Engineer would find challenging but fair, reflecting real-world scenarios they encounter in enterprise environments.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment