AWS Services, Explained in Five Questions
Every service card answers the questions that actually matter when you meet a new AWS service: What is it, why does it exist, when should you pick it, where does it sit in an architecture, and how do you integrate it — plus exam-angle tips and links to the official AWS documentation and blogs. Coverage is tuned to the AWS Certified AI Practitioner (AIF-C01) exam.
Show services for your learning path
The 16 services below are the ones the AWS Machine Learning Associate path expects you to know. Open the study path →
✨ Generative AI · 2
Amazon BedrockFully managed access to foundation models through one API.A serverless service that exposes foundation models from Amazon (Nova, Titan), Anthropic (Claude), Meta (Llama), Mistral, and others through a sing…What · Why · When · Where · How →Amazon SageMaker JumpStartA model hub for deploying and fine-tuning pretrained models in your account.A catalogue of hundreds of open-weight and proprietary foundation and task models (Llama, Mistral, Falcon, Stable Diffusion, plus classic vision/NL…What · Why · When · Where · How →
🧠 ML Platform (SageMaker) · 7
Amazon SageMaker AIThe end-to-end platform for building, training, and deploying ML models.AWS's managed machine-learning platform covering the whole lifecycle: notebooks (Studio), data prep, distributed training, hyperparameter tuning, m…What · Why · When · Where · How →Amazon SageMaker ClarifyBias detection and model explainability across the ML lifecycle.A SageMaker capability that measures statistical bias in datasets (pre-training) and models (post-training), explains individual predictions using …What · Why · When · Where · How →Amazon SageMaker Data WranglerVisual, low-code data preparation for machine learning.A visual interface in SageMaker Studio for importing, exploring, transforming, and featurising data with 300+ built-in transformations — turning wh…What · Why · When · Where · How →Amazon SageMaker Ground TruthHuman labeling workforces and workflows for training data.A data-labeling service that routes images, text, and video to human annotators — your own team, third-party vendors, or the Mechanical Turk crowd …What · Why · When · Where · How →Amazon SageMaker CanvasNo-code machine learning for business analysts.A point-and-click application where non-programmers build, evaluate, and use ML models — tabular prediction, forecasting, and even fine-tuning foun…What · Why · When · Where · How →Amazon SageMaker Model MonitorContinuous monitoring of deployed models for drift and quality.A monitoring capability that captures endpoint traffic and periodically checks it against a training-time baseline for data-quality drift, model-qu…What · Why · When · Where · How →Amazon SageMaker Feature StoreA central, consistent repository for ML features.A purpose-built store for ML features with an online store (low-latency reads at inference) and an offline store (historical data in S3 for trainin…What · Why · When · Where · How →
🤖 AI Services · 1
🗄️ Data & Analytics · 4
Amazon S3Object storage — the data backbone of every AI workload.Durable (11 nines), effectively unlimited object storage organised in buckets, with storage classes from frequent-access to deep archive, versionin…What · Why · When · Where · How →AWS GlueServerless ETL and the Data Catalog for your lake.A serverless data-integration service: crawlers discover schemas into a central Data Catalog, Spark/Python jobs transform data at scale, and DataBr…What · Why · When · Where · How →Amazon AthenaServerless SQL directly over data in S3.An interactive query service that runs standard SQL against files in S3 (CSV, JSON, Parquet, ORC) using the Glue Data Catalog for schemas — no clus…What · Why · When · Where · How →Amazon OpenSearch ServiceSearch, log analytics, and vector database for RAG.A managed OpenSearch cluster (and serverless option) for full-text search and log analytics that now doubles as a vector database: k-NN indexes sto…What · Why · When · Where · How →
⚙️ Compute & Cost · 2
Amazon EC2 Accelerated InstancesGPU and AI-accelerator instances for training and inference.EC2 instance families with hardware accelerators: P-series (NVIDIA GPUs for training), G-series (graphics/inference GPUs), and Trn/Inf series built…What · Why · When · Where · How →AWS Trainium & InferentiaAWS-designed silicon for cheaper training and inference.Purpose-built machine-learning chips: Trainium (Trn instances) accelerates model TRAINING; Inferentia (Inf instances) accelerates INFERENCE — both …What · Why · When · Where · How →
Preparing for the exam? Start the AI Practitioner study guide or jump into the AIF-C01 learning path.