Skip to content

khasky/aws-ai-practitioner

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

2 Commits
 
 
 
 
 
 
 
 

Repository files navigation

Everything you need to know to become AWS AI Practitioner

This repo is a study guide for the AWS Certified AI Practitioner (AIF-C01) exam. It also covers how common AI and generative AI ideas show up on AWS in real projects.

The material follows the official exam guide (PDF, HTML hub). For exam scope, service names, and domain weights, rely on AWS docs and the guide; both can change.


Who this is for

  • Professionals who need foundational AI/ML and generative AI literacy on AWS.
  • Candidates preparing for AIF-C01 who want domain-by-domain coverage plus hands-on examples.
  • Anyone mapping business problems to the right AWS AI building blocks (without assuming you will train models from scratch).

The exam’s target candidate uses AI/ML on AWS but is not expected to implement deep model engineering, heavy MLOps pipelines, or organization-wide governance frameworks. On the exam those topics show up as concepts to recognize, not as tasks to perform.


Exam snapshot (official basics)

Item Detail
Exam code AIF-C01
Level Foundational (AWS Certification)
Question types Multiple choice, multiple response, ordering, matching
Scored questions 50 (plus 15 unscored questions that do not affect your score)
Passing score 700 on a scaled score of 100–1000
Scoring model Compensatory (overall pass; section weights differ)

Recommended knowledge (from AWS): familiarity with core AWS services (for example EC2, S3, Lambda, Amazon Bedrock, Amazon SageMaker AI), the shared responsibility model, IAM, and pricing models. Up to about six months’ exposure to AI/ML on AWS is typical for the target candidate.

Out-of-scope job tasks (examples from AWS): developing model algorithms, heavy feature engineering, hyperparameter tuning, building full AI/ML pipelines or security/compliance programs. On the exam you need to recognize what these are, not perform them at expert depth.


Content domains and weights (scored content)

Domain Topic Weight
1 Fundamentals of AI and ML 20%
2 Fundamentals of GenAI 24%
3 Applications of Foundation Models 28%
4 Guidelines for Responsible AI 14%
5 Security, Compliance, and Governance for AI Solutions 14%

How to use this repo

  1. Read the domain guides in order (1 through 5). Later domains assume vocabulary from earlier ones.
  2. Cross-link to AWS docs for anything operational (IAM, encryption, regional availability, pricing).
  3. Run the code examples under examples/ to connect API shapes to the concepts (Bedrock, boto3 patterns, evaluation and monitoring ideas).
  4. Validate exam scope using the official in-scope services list.

Suggested study sequence (example)

Phase Focus Activities
1 Vocabulary & lifecycle Domain 1 guide; sketch one ML lifecycle for a business problem you know.
2 GenAI building blocks Domain 2 guide; list 3 GenAI use cases and 2 failure modes (hallucination, cost).
3 FMs in production Domain 3 guide; practice explaining RAG, agents, and evaluation metrics out loud.
4 Responsibility & trust Domain 4 guide; map tools (Guardrails, Clarify, Model Monitor, A2I) to risks.
5 Security & governance Domain 5 guide; trace IAM → encryption → logging for a Bedrock workload on paper.
6 Service mapping guides/06-aws-services-primer.md; drill “which service for which scenario?”
7 Hands-on Run Bedrock examples in a sandbox account; adjust inference parameters and observe changes.

Guide index

Guide File
Domain 1: AI & ML fundamentals guides/01-fundamentals-ai-and-ml.md
Domain 2: Generative AI fundamentals guides/02-fundamentals-of-genai.md
Domain 3: Foundation models in applications guides/03-applications-of-foundation-models.md
Domain 4: Responsible AI guides/04-responsible-ai.md
Domain 5: Security, compliance, governance guides/05-security-compliance-governance.md
AWS services at a glance (exam-oriented) guides/06-aws-services-primer.md

Code examples

Example Description
examples/bedrock_converse.py Invoke a foundation model with the Converse API (messages, inference parameters).
examples/bedrock_embeddings.py Generate embeddings for RAG-style workflows.
examples/rag_similarity_concept.py Cosine similarity between vectors (RAG retrieval concept).
examples/requirements.txt Minimal Python dependencies for the samples.

Examples assume credentials via the default AWS credential chain (for example environment variables, ~/.aws/credentials, or an IAM role). Replace model IDs and regions with values valid for your account.


Official resources


Disclaimer

This guide is educational and not affiliated with AWS. Exams, service names, and guides change; check AWS documentation before you schedule a test or design production systems.

Contributors