AI-900 Cheat Sheet

AI-900 Exam Blueprint — At a Glance

#	Domain	Weight	Questions	Practice
1.0	Describe Artificial Intelligence workloads and considerations Describe Artificial Intelligence workloads and considerations questions on this certification test your ability to deploy and manage describe artificial intelligence workloads and considerations concepts in scenario-based situations.	—	199	Practice →
2.0	Describe fundamental principles of machine learning on Azure Describe fundamental principles of machine learning on Azure questions on this certification test your ability to deploy and manage describe fundamental principles of machine learning on azure concepts in scenario-based situations.	—	207	Practice →
3.0	Describe features of computer vision workloads on Azure Describe features of computer vision workloads on Azure questions on this certification test your ability to deploy and manage describe features of computer vision workloads on azure concepts in scenario-based situations.	—	208	Practice →
4.0	Describe features of Natural Language Processing workloads on Azure NAT questions usually test how private addresses are translated, when to use static NAT, dynamic NAT or PAT, and how inside/outside interfaces affect traffic flow.	—	200	Practice →
5.0	Describe features of generative AI workloads on Azure Describe features of generative AI workloads on Azure questions on this certification test your ability to deploy and manage describe features of generative ai workloads on azure concepts in scenario-based situations.	—	206	Practice →

Domain Quick Reference

1.0Describe Artificial Intelligence workloads and considerations

Describe Artificial Intelligence workloads and considerations questions on this certification test your ability to deploy and manage describe artificial intelligence workloads and considerations concepts in scenario-based situations.

Key concepts

✓Core Describe Artificial Intelligence workloads and considerations concepts and how they apply in real-world cloud scenarios.
✓How to deploy describe artificial intelligence workloads and considerations correctly and verify the outcome.
✓Troubleshooting describe artificial intelligence workloads and considerations issues by interpreting error output and system state.
✓Cloud best practices and Describe Artificial Intelligence workloads and considerations design trade-offs tested by this certification.

Watch out for

⚠Selecting the most expensive service when a simpler managed option meets the requirement.
⚠Forgetting that cloud resources must be explicitly secured — defaults are rarely secure.
⚠Choosing a global service fix when the issue is region-specific.
⚠Overlooking cost implications of cross-region data transfer in architecture questions.

2.0Describe fundamental principles of machine learning on Azure

Describe fundamental principles of machine learning on Azure questions on this certification test your ability to deploy and manage describe fundamental principles of machine learning on azure concepts in scenario-based situations.

Key concepts

✓Core Describe fundamental principles of machine learning on Azure concepts and how they apply in real-world cloud scenarios.
✓How to deploy describe fundamental principles of machine learning on azure correctly and verify the outcome.
✓Troubleshooting describe fundamental principles of machine learning on azure issues by interpreting error output and system state.
✓Cloud best practices and Describe fundamental principles of machine learning on Azure design trade-offs tested by this certification.

Watch out for

⚠Selecting the most expensive service when a simpler managed option meets the requirement.
⚠Forgetting that cloud resources must be explicitly secured — defaults are rarely secure.
⚠Choosing a global service fix when the issue is region-specific.
⚠Overlooking cost implications of cross-region data transfer in architecture questions.

3.0Describe features of computer vision workloads on Azure

Describe features of computer vision workloads on Azure questions on this certification test your ability to deploy and manage describe features of computer vision workloads on azure concepts in scenario-based situations.

Key concepts

✓Core Describe features of computer vision workloads on Azure concepts and how they apply in real-world cloud scenarios.
✓How to deploy describe features of computer vision workloads on azure correctly and verify the outcome.
✓Troubleshooting describe features of computer vision workloads on azure issues by interpreting error output and system state.
✓Cloud best practices and Describe features of computer vision workloads on Azure design trade-offs tested by this certification.

Watch out for

⚠Selecting the most expensive service when a simpler managed option meets the requirement.
⚠Forgetting that cloud resources must be explicitly secured — defaults are rarely secure.
⚠Choosing a global service fix when the issue is region-specific.
⚠Overlooking cost implications of cross-region data transfer in architecture questions.

4.0Describe features of Natural Language Processing workloads on Azure

NAT questions usually test how private addresses are translated, when to use static NAT, dynamic NAT or PAT, and how inside/outside interfaces affect traffic flow.

Key concepts

✓Static NAT, dynamic NAT and PAT behaviour.
✓Inside local, inside global, outside local and outside global address meanings.
✓How NAT affects connectivity between private and public destinations.
✓How to troubleshoot NAT rules, ACL matches and interface direction.

Watch out for

⚠PAT allows many inside hosts to share one public address by using port numbers.
⚠NAT rules depend on correct inside and outside interface configuration.
⚠The ACL used for NAT identifies traffic to translate; it is not a security ACL.
⚠Static NAT maps one private address to one public address; PAT overloads translations.

5.0Describe features of generative AI workloads on Azure

Describe features of generative AI workloads on Azure questions on this certification test your ability to deploy and manage describe features of generative ai workloads on azure concepts in scenario-based situations.

Key concepts

✓Core Describe features of generative AI workloads on Azure concepts and how they apply in real-world cloud scenarios.
✓How to deploy describe features of generative ai workloads on azure correctly and verify the outcome.
✓Troubleshooting describe features of generative ai workloads on azure issues by interpreting error output and system state.
✓Cloud best practices and Describe features of generative AI workloads on Azure design trade-offs tested by this certification.

Watch out for

⚠Selecting the most expensive service when a simpler managed option meets the requirement.
⚠Forgetting that cloud resources must be explicitly secured — defaults are rarely secure.
⚠Choosing a global service fix when the issue is region-specific.
⚠Overlooking cost implications of cross-region data transfer in architecture questions.

Exam Day Reminders

→Read every question stem fully — look for qualifiers like 'MOST likely,' 'BEST,' or 'EXCEPT.'
→Flag uncertain questions and come back — don't waste time on one question.
→Eliminate obviously wrong options first, then choose between remaining ones.
→Trust your first instinct unless you have a specific reason to change.
→For AI-900, scenarios typically have one clearly best answer — look for the option that matches the specific constraints in the question.

Domain

Weight

Questions

Practice

1.0

Describe Artificial Intelligence workloads and considerations

—

199

Practice →

2.0

Describe fundamental principles of machine learning on Azure

—

207

Practice →

3.0

Describe features of computer vision workloads on Azure

—

208

Practice →

4.0

Describe features of Natural Language Processing workloads on Azure

NAT questions usually test how private addresses are translated, when to use static NAT, dynamic NAT or PAT, and how inside/outside interfaces affect traffic flow.

—

200

Practice →

5.0

Describe features of generative AI workloads on Azure

—

206

Practice →

Domain Quick Reference

1.0Describe Artificial Intelligence workloads and considerations

Key concepts

✓Core Describe Artificial Intelligence workloads and considerations concepts and how they apply in real-world cloud scenarios.
✓How to deploy describe artificial intelligence workloads and considerations correctly and verify the outcome.
✓Troubleshooting describe artificial intelligence workloads and considerations issues by interpreting error output and system state.
✓Cloud best practices and Describe Artificial Intelligence workloads and considerations design trade-offs tested by this certification.

Watch out for

⚠Selecting the most expensive service when a simpler managed option meets the requirement.
⚠Forgetting that cloud resources must be explicitly secured — defaults are rarely secure.
⚠Choosing a global service fix when the issue is region-specific.
⚠Overlooking cost implications of cross-region data transfer in architecture questions.

2.0Describe fundamental principles of machine learning on Azure

Key concepts

✓Core Describe fundamental principles of machine learning on Azure concepts and how they apply in real-world cloud scenarios.
✓How to deploy describe fundamental principles of machine learning on azure correctly and verify the outcome.
✓Troubleshooting describe fundamental principles of machine learning on azure issues by interpreting error output and system state.
✓Cloud best practices and Describe fundamental principles of machine learning on Azure design trade-offs tested by this certification.

Watch out for

⚠Selecting the most expensive service when a simpler managed option meets the requirement.
⚠Forgetting that cloud resources must be explicitly secured — defaults are rarely secure.
⚠Choosing a global service fix when the issue is region-specific.
⚠Overlooking cost implications of cross-region data transfer in architecture questions.

3.0Describe features of computer vision workloads on Azure

Key concepts

✓Core Describe features of computer vision workloads on Azure concepts and how they apply in real-world cloud scenarios.
✓How to deploy describe features of computer vision workloads on azure correctly and verify the outcome.
✓Troubleshooting describe features of computer vision workloads on azure issues by interpreting error output and system state.
✓Cloud best practices and Describe features of computer vision workloads on Azure design trade-offs tested by this certification.

Watch out for

⚠Selecting the most expensive service when a simpler managed option meets the requirement.
⚠Forgetting that cloud resources must be explicitly secured — defaults are rarely secure.
⚠Choosing a global service fix when the issue is region-specific.
⚠Overlooking cost implications of cross-region data transfer in architecture questions.

4.0Describe features of Natural Language Processing workloads on Azure

NAT questions usually test how private addresses are translated, when to use static NAT, dynamic NAT or PAT, and how inside/outside interfaces affect traffic flow.

Key concepts

✓Static NAT, dynamic NAT and PAT behaviour.
✓Inside local, inside global, outside local and outside global address meanings.
✓How NAT affects connectivity between private and public destinations.
✓How to troubleshoot NAT rules, ACL matches and interface direction.

Watch out for

⚠PAT allows many inside hosts to share one public address by using port numbers.
⚠NAT rules depend on correct inside and outside interface configuration.
⚠The ACL used for NAT identifies traffic to translate; it is not a security ACL.
⚠Static NAT maps one private address to one public address; PAT overloads translations.

5.0Describe features of generative AI workloads on Azure

Key concepts

✓Core Describe features of generative AI workloads on Azure concepts and how they apply in real-world cloud scenarios.
✓How to deploy describe features of generative ai workloads on azure correctly and verify the outcome.
✓Troubleshooting describe features of generative ai workloads on azure issues by interpreting error output and system state.
✓Cloud best practices and Describe features of generative AI workloads on Azure design trade-offs tested by this certification.

Watch out for

⚠Selecting the most expensive service when a simpler managed option meets the requirement.
⚠Forgetting that cloud resources must be explicitly secured — defaults are rarely secure.
⚠Choosing a global service fix when the issue is region-specific.
⚠Overlooking cost implications of cross-region data transfer in architecture questions.

Exam Day Reminders

→Read every question stem fully — look for qualifiers like 'MOST likely,' 'BEST,' or 'EXCEPT.'

→Flag uncertain questions and come back — don't waste time on one question.

→Eliminate obviously wrong options first, then choose between remaining ones.

→Trust your first instinct unless you have a specific reason to change.

→For AI-900, scenarios typically have one clearly best answer — look for the option that matches the specific constraints in the question.

AI-900 Exam Blueprint — At a Glance

Domain Quick Reference

1.0Describe Artificial Intelligence workloads and considerations

2.0Describe fundamental principles of machine learning on Azure

3.0Describe features of computer vision workloads on Azure

4.0Describe features of Natural Language Processing workloads on Azure

5.0Describe features of generative AI workloads on Azure

Exam Day Reminders

More AI-900 resources

AI-900 Cheat Sheet

AI-900 Exam Blueprint — At a Glance

Domain Quick Reference

1.0Describe Artificial Intelligence workloads and considerations

2.0Describe fundamental principles of machine learning on Azure

3.0Describe features of computer vision workloads on Azure

4.0Describe features of Natural Language Processing workloads on Azure

5.0Describe features of generative AI workloads on Azure

Exam Day Reminders

More AI-900 resources