AWS AI/ML Services Overview
AWS offers AI/ML a large number of services across three tiers. Here's a breakdown of the key services available at each level.
AI Services
Pre-built models that work immediately. No ML knowledge required.
| Service | What It Does |
|---|---|
| Amazon Comprehend | Extract insights from text (sentiment, key phrases, language) |
| Amazon Polly | Convert text to lifelike speech in multiple languages |
| Amazon Transcribe | Convert speech to text with speaker identification |
| Amazon Translate | Translate text between languages in real-time |
| Amazon Kendra | Intelligent search that understands natural language |
| Amazon Rekognition | Analyze images and videos (objects, faces, text, scenes) |
| Amazon Textract | Extract text from documents, forms, and tables |
| Amazon Lex | Build chatbots with voice and text interfaces |
| Amazon Personalize | Create personalized recommendations from historical data |
ML Services
Build, train, and deploy your own ML models. AWS manages the infrastructure.
| Service | What It Does |
|---|---|
| Amazon SageMaker AI | Full ML platform with IDE for data scientists, no-code interface for business analysts, managed infrastructure, and automated MLOps workflows |
ML Frameworks and Infrastructure
Full control to build custom models from scratch. Requires ML expertise.
| Service | What It Does |
|---|---|
| ML Frameworks | Pre-built libraries (PyTorch, TensorFlow) for building custom models |
| ML Infrastructure | ML-optimized compute instances (EC2, EMR, ECS) for training and inference |
