AWS AI Applications: Types, Services, and APIs Guide

Table of Contents

• Overview
• AWS AI Service Landscape
• Application Categories
  • 1. Conversational AI Applications
  • 2. Content Generation Applications
  • 3. Document Processing Applications
  • 4. Search and Discovery Applications
  • 5. Data Analysis Applications
  • 6. Automation Applications
• Agentic Applications
  • What are Agentic Applications?
  • Agentic Application Categories
  • Building Agentic Applications
• Service and API Mapping
• Architecture Patterns
• Best Practices

---

Overview

AWS provides a comprehensive suite of AI services spanning from pre-trained AI services to foundation models and custom ML. This guide maps application types to the appropriate AWS services and APIs.

AWS AI Stack:

┌─────────────────────────────────────────────────────────────┐
│                    APPLICATION LAYER                         │
│  Your AI-powered applications and solutions                 │
└─────────────────────────────────────────────────────────────┘
                            ↓
┌─────────────────────────────────────────────────────────────┐
│                   GENERATIVE AI LAYER                        │
│  • Amazon Bedrock (Foundation Models)                       │
│  • Amazon Q (Business AI Assistant)                         │
│  • Amazon CodeWhisperer (Code Generation)                   │
└─────────────────────────────────────────────────────────────┘
                            ↓
┌─────────────────────────────────────────────────────────────┐
│                   AI SERVICES LAYER                          │
│  • Amazon Comprehend (NLP)                                  │
│  • Amazon Textract (Document Analysis)                      │
│  • Amazon Rekognition (Computer Vision)                     │
│  • Amazon Polly (Text-to-Speech)                           │
│  • Amazon Transcribe (Speech-to-Text)                       │
│  • Amazon Translate (Language Translation)                  │
│  • Amazon Lex (Conversational AI)                          │
│  • Amazon Personalize (Recommendations)                     │
└─────────────────────────────────────────────────────────────┘
                            ↓
┌─────────────────────────────────────────────────────────────┐
│                   ML PLATFORM LAYER                          │
│  • Amazon SageMaker (Custom ML)                             │
│  • AWS Deep Learning AMIs                                   │
│  • AWS Deep Learning Containers                             │
└─────────────────────────────────────────────────────────────┘

AWS AI Service Landscape

Generative AI Services

Service	Purpose	Primary API
Amazon Bedrock	Foundation models, RAG, Agents	bedrock-runtime, bedrock-agent-runtime
Amazon Q	Business AI assistant	q-business, q-connect
Amazon CodeWhisperer	Code generation	IDE integration, codewhisperer

AI/ML Services

Service	Purpose	Primary API
Amazon Comprehend	NLP, sentiment, entities	comprehend
Amazon Textract	Document text extraction	textract
Amazon Rekognition	Image/video analysis	rekognition
Amazon Polly	Text-to-speech	polly
Amazon Transcribe	Speech-to-text	transcribe
Amazon Translate	Language translation	translate
Amazon Lex	Chatbots	lexv2-runtime
Amazon Personalize	Recommendations	personalize-runtime
Amazon Forecast	Time series forecasting	forecast
Amazon Kendra	Intelligent search	kendra

ML Platform

Service	Purpose	Primary API
Amazon SageMaker	Custom ML models	sagemaker, sagemaker-runtime