Developing Generative AI Applications on AWS

Module 1: Exploring Generative AI Application Components

• Generative AI fundamentals
• Machine learning and generative AI
• Foundation models
• Large language models
• Generative AI application architecture
• AWS generative AI technology stack
• Components of GenAI applications
• Common generative AI use cases
• Business applications of GenAI
• Risks and benefits

Module 2: Programming with Amazon Bedrock

• Introduction to Amazon Bedrock
• Amazon Bedrock architecture
• Foundation model access
• Bedrock APIs
• AWS SDK integration
• Programmatic model invocation
• Inference parameters
• Controlling model responses
• Streaming responses
• Error handling
• Application integration

Hands-On Labs:

• Develop with Amazon Bedrock APIs
• Develop streaming patterns with Amazon Bedrock APIs

Module 3: Prompt Engineering for Developers

• Prompt engineering fundamentals
• Prompt structure
• Instruction design
• Context and constraints
• Few-shot prompting
• Zero-shot prompting
• Chain-of-thought considerations
• Prompt optimization
• Model-specific prompting
• Prompt testing
• Improving response quality

Module 4: Amazon Bedrock APIs in Application Architectures

• Common Bedrock application patterns
• Integrating foundation models into applications
• Conversational AI
• Chat applications
• Managing conversational context
• Conversation memory
• API-based GenAI architectures
• Application workflows
• Error handling and resilience

Hands-On Lab:

• Develop conversation patterns with Amazon Bedrock APIs

Module 5: Retrieval Augmented Generation

• RAG fundamentals
• RAG architecture
• Embeddings
• Vector search
• Knowledge retrieval
• Document ingestion
• Chunking concepts
• Amazon Bedrock Knowledge Bases
• Retrieval and generation
• Grounding model responses
• RAG application development
• RAG evaluation

Hands-On Lab:

• Develop RAG applications using Amazon Bedrock Knowledge Bases

Module 6: Customizing Generative AI Applications

• Foundation model customization
• Model selection
• Fine-tuning concepts
• Retrieval-based customization
• Prompt optimization
• Application-specific context
• Knowledge integration
• Model response evaluation
• Customization trade-offs

Module 7: Securing Generative AI Applications

• Generative AI security fundamentals
• AWS IAM
• Authentication and authorization
• Data protection
• Encryption
• Secure API access
• Data privacy
• Prompt security
• Application security
• Responsible AI considerations
• Guardrails and content controls

Module 8: Building Production-Ready GenAI Applications

• Application scalability
• Performance optimization
• Cost considerations
• Model selection
• Inference optimization
• Monitoring
• Logging
• Error handling
• Application resilience
• Responsible AI
• Production deployment considerations

Module 9: Generative AI Application Project

• Define a GenAI application use case
• Select an appropriate foundation model
• Design the application architecture
• Integrate Amazon Bedrock
• Develop prompts
• Implement conversational functionality
• Build a RAG workflow
• Connect a knowledge base
• Apply security controls
• Test model responses
• Evaluate application performance
• Optimize the solution

• Software Developers
• Generative AI Developers
• AI Engineers
• Machine Learning Engineers
• Cloud Developers
• Application Developers
• Full-Stack Developers
• Backend Developers
• AWS Developers
• Solutions Architects
• Cloud Engineers
• MLOps Engineers
• Developers building LLM-powered applications
• Professionals developing RAG-based applications
• Professionals preparing for advanced AWS Generative AI certifications

• Understand the architecture of generative AI applications.
• Work programmatically with Amazon Bedrock APIs.
• Integrate foundation models into cloud applications.
• Develop effective prompt engineering strategies.
• Optimize prompts for improved model responses.
• Build conversational AI applications.
• Implement conversational memory and context.
• Understand embeddings and vector-based retrieval.
• Build Retrieval Augmented Generation (RAG) applications.
• Use Amazon Bedrock Knowledge Bases.
• Understand foundation model customization and fine-tuning concepts.
• Apply security principles to generative AI applications.
• Design scalable and production-oriented GenAI architectures.
• Evaluate and optimize generative AI application performance and cost.
• Build practical generative AI applications using AWS services.

Required

• Completion of Generative AI Essentials on AWS or equivalent knowledge.
• Intermediate-level proficiency in Python.
• Familiarity with AWS Cloud.

Recommended

• Basic understanding of machine learning and generative AI.
• Familiarity with APIs and application development.
• Understanding of cloud computing concepts.
• Basic knowledge of AWS IAM and security.
• Familiarity with databases and data storage is beneficial.
• Understanding of software development practices.

Course: Developing Generative AI Applications on AWS
Level: Advanced
Duration: 2 Days
Delivery: Instructor-led training
Hands-On: Yes
Technology: Amazon Web Services
Primary Service: Amazon Bedrock
Focus: Generative AI application development, foundation models, prompt engineering, RAG, and application architecture