As the famous quote goes, AI might not take your job, but the person with AI knowledge will.
🎥 Video link
Yay! I've cleared the Amazon Web Services Certified AI Practitioner Beta Exam. This marks my sixth Amazon Web Services certification and my thirty-third overall.
⚠️ WARNING: Some Housekeeping Tasks Before Reading This Blog or Watching the Video
🛑 Since everyone needs to sign an NDA with Amazon Web Services, I can't share the exact questions asked during the exam, nor do I have a photographic memory. However, I can provide some pointers on what to expect in the exam.
🔄 As we all know, Amazon Web Services infrastructure updates frequently, so some information might become outdated after a few days, weeks, or months.
🚫 Please don't ask for any exam dumps or questions; that defeats the whole purpose of taking the exam.
📊 Exam Stats
Duration: 120 minutes
Number of Questions: 85 (for beta)
📚 Exam Preparation
You don't need to spend any money to prepare for this exam, but you should be aware of a few key services that I'll cover later. I used official Amazon Web Services resources to prepare for the exam:
https://skillbuilder.aws/exam-prep/ai-practitioner
Based on my experience, you must be familiar with these five services to clear the exam:
✅ Amazon Bedrock: Nearly half of the exam covers Bedrock-related questions. To pass, you must know this service. Make sure you understand Guardrails, Knowledge Base, Agents, Model Evaluation, etc.
✅ Amazon SageMaker: This is the second most asked service, and you must know it well. Be familiar with SageMaker Studio, Canvas, Ground Truth, JumpStart, and others.
✅ Machine Learning Terminology: You should know key machine learning terms like Epoch, MSE, RMSE, etc.
✅ Generative AI Terminology: Understanding Generative AI terminology, such as Top-P, temperature, hallucination, tokens, vector embeddings, etc., is essential.
✅ Amazon Web Services Core Services: They assume you're comfortable with core Amazon Web Services services like Identity and Access Management (IAM), CloudTrail, S3, etc.
👎 Things I Didn't Like About the Exam
❌ Focus on Older Services: Amazon Web Services has several machine learning services from before the Generative AI era that are now considered outdated, like Comprehend, Polly, and Rekognition, yet they are still heavily featured in the exam. Some of these services have confusing names that don't make much sense.
❌ Complexity of Bedrock and SageMaker: Amazon Bedrock and SageMaker are not single services but rather collections of services, which can be especially challenging for newcomers to remember, particularly with SageMaker.
⏪ Before you write the exam, please take advantage of the sample exams provided by Amazon Web Services. These will gauge your knowledge and help you understand where you stand in terms of preparation. Official Practice Question Set - Amazon Web Services Certified AI Practitioner
💡Whether you plan to take this exam or not, knowledge of AWS Bedrock will be essential in the coming year if you use AWS. This service is set to become a de facto standard, and everyone should be as familiar with it as they are with EC2 and S3.
📝 Final Words
Most of the questions are straightforward, with few scenario-based questions, meaning less reading. Stay calm, take the exam, and feel free to reach out if you have any questions.
Top comments (0)