Preparing for the AWS Certified AI Practitioner (AIF-C01) can feel overwhelming — not because the concepts are complex, but because the exam covers a wide range of AI terminology, AWS services, ML workflows, prompt engineering, RAG, and evaluation metrics.

When I started preparing for the AWS Certified AI Practitioner (AIF-C01) exam, I quickly realized something — the content wasn’t “hard,” but there was so much to remember, and many terms sounded similar:

• Supervised vs Unsupervised
• Evaluation metrics
• SageMaker services
• Bedrock features
• Prompt engineering techniques
• RAG components

To keep things simple, I began writing down small memory hooks, short patterns, and mental shortcuts on a notepad.

These hooks helped me instantly recall concepts during the exam — especially when faced with confusingly worded scenario questions.

During my preparation, I followed the excellent QA/CloudAcademy course:“AWS Certified AI Practitioner (AIF-C01) Certification Preparation” by Danny Jessee.

This course helped me understand how different services fit together, while my memory hooks helped me recall the details under exam pressure.

This blog post is a summary of all the memory hooks that helped me pass the exam — shared so you (and others) can benefit as well.

To make learning efficient, here is a single consolidated guide of the best memory hooks and mnemonics used to successfully pass the exam — now shared so others can benefit too.

🧠 1. Machine Learning Basics

Supervised vs Unsupervised

Labels → Supervised No Labels → Unsupervised

✔ Supervised = Teacher + Correct Answers ✔ Unsupervised = Find patterns (clustering, segments)

Classification vs Regression

Classes → Classification Numbers → Regression

Overfitting vs Underfitting

Overfitting = Too complex → Increase regularization Underfitting = Too simple → Decrease regularization

🧠 2. Key Algorithms

Clustering

Group customers? No labels? → K-Means

Image Classification

Flower Classification → k-NN or Decision Tree

Anomaly Detection

No labels + abnormal detection → Autoencoders

🧠 3. GenAI Prompt Engineering

• Few-shot prompting
• Show format → Few-shot prompting.
• Prompt chaining
• Multi-step workflow → Prompt chaining.
• ReAct prompting

Reason + Action + Tool use → ReAct.

Temperature

Creativity ↑ → Temperature ↑ Consistency ↑ → Temperature ↓

🧠 4. LLM Inference Parameters

• Temperature: Creativity
• Top-K: Number of token choices
• Top-P: Probability bucket
• Max Tokens: Output length
• Frequency Penalty: Reduce repeated words
• Presence Penalty: Discourage repeated topics

Creativity → Temp / Top-K / Top-P Length → Max Tokens Repetition → Frequency & Presence

🧠 5. RAG (Retrieval-Augmented Generation)

Purpose of Chunking

Chunking = Better retrieval → Better context

Batch Steps in RAG

✔ Content embeddings ✔ Build search index (NOT query embeddings or response generation)

LLM Type for Multimodal Search

Text + Image queries → Multimodal model

🧠 6. Evaluating ML Models

Summarization Metrics

Summarization → ROUGE (If ROUGE missing → Choose BLEU)

Translation Metrics

Translation → BLEU / METEOR Classification Metrics Imbalanced data → F1 Score Balanced → Accuracy

Regression Metrics

• Numeric prediction → MSE / RMSE
• LLM Quality

Perplexity → How surprised is the model?

🧠 7. AWS Services — Quick Memory Hooks Model Cards

Governance + Documentation → Model Cards

Model Monitor

Detect drift in production → Model Monitor

Ground Truth

Human labeling → Ground Truth

JumpStart

Pre-built models + quick deploy → JumpStart

SageMaker Canvas

No-code data prep → Canvas

HealthScribe

Medical speech-to-text → HealthScribe

Guardrails for Bedrock

Responsible AI (safety filters) → Guardrails

PartyRock

Experiment + Learn + No cost → PartyRock (Not for VPC, not for deployments)

🧠 8. GenAI Lifecycle

Design → Data → Train → Evaluate → Deploy → Monitor

Evaluation Stage

Accuracy testing

Safety + toxicity testing

Hallucination measurements

Inference

• Train = Learn
• Infer = Predict
• Deploy = Serve

🧠 9. Embeddings

Embeddings = Meaning → Vectors Reduced dimension → Same meaning → Similarity search

🧠 10. Foundational Concepts

Fine-tuning

Teach big model a small task well. ✔ Domain-specific labeled data ✔ Improves specific task performance ✔ NOT retraining from scratch ✔ NOT updating model to recent events

Responsible AI

Safety + Filters + Detect toxicity → Use Guardrails

🎉 Final Thoughts

These memory hooks are designed with one purpose:

👉 Make recall instant during the exam 👉 Reduce confusion between similar concepts 👉 Build confidence with patterns instead of memorising definitions