Introduction of AWS Managed AI Services
Introduction to AWS managed AI services for AIF-C01: why to use pre-trained AI services, and an overview of the main services for text, speech and vision.
DashboardIntroduction of AWS Managed AI Services
Why Use AWS AI Managed Services?
- AWS Managed AI Services are pre-trained machine learning services designed to solve specific use cases without requiring you to build or train models from scratch.
- Using managed services helps you focus on your application logic rather than infrastructure, training, or scaling ML workloads.
- You can run machine learning manually on your own servers, but AWS managed services provide major advantages.
- Responsiveness & High Availability
- Services are available across multiple AWS Regions.
- Built to be highly responsive even under heavy load.
- Redundancy & Reliability
- Deployed across multiple Availability Zones (AZs).
- Keeps running even if one AZ experiences a failure.
- High Performance
- Backed by optimized CPUs, GPUs, and accelerators.
- Provides better performance and lower cost vs. building everything manually.
- Token-Based Pricing
- You pay only for what you use (per request or per token).
- No need to over-provision or operate servers.
Examples of AWS AI Managed Services
- Generative AI
- Amazon Bedrock – Foundation models and GenAI apps
- Amazon Q Business / Q Developer – Higher-level GenAI assistants
- Text & Document Processing
- Amazon Comprehend – NLP, sentiment, key phrases
- Amazon Translate – Language translation
- Amazon Textract – Document OCR and extraction
- Vision
- Amazon Rekognition – Image and video analysis
- Search
- Amazon Kendra – Intelligent enterprise search
- Chatbots
- Amazon Lex – Conversational interfaces
- Speech
- Amazon Polly – Text to speech
- Amazon Transcribe – Speech to text
- Recommendations
- Amazon Personalize – Real-time recommendations
- Full ML Platform
- Amazon SageMaker – Build, train, tune, deploy, and monitor ML models