Amazon Bedrock & Foundation Modelsin Bengaluru

Module 1: Introduction to Amazon Bedrock & Foundation Models

• Overview of Generative AI and foundation models
  • Understanding Large Language Models (LLMs)
  • Types and capabilities of foundation models
• Introduction to Amazon Bedrock
  • Key features and supported AI capabilities
  • Business applications of foundation models
• Industry use cases and enterprise AI adoption
• Introduction to Amazon Bedrock training in Bengaluru, India

Module 2: AWS Environment Setup

• Introduction to the AWS Management Console
  • AWS accounts, regions, and service access
  • Understanding IAM roles and permissions
• Accessing Amazon Bedrock
  • Service configuration and model availability
  • Understanding model access requirements
• Setting up the development environment
  • Python installation and package management
  • AWS CLI and SDK configuration

Module 3: Foundation Models and Model Selection

• Understanding foundation model architectures and capabilities
  • Text generation and conversational models
  • Embedding and multimodal model concepts
• Exploring supported models in Amazon Bedrock
  • Comparing model capabilities and use cases
  • Selecting models based on application requirements
• Understanding model limitations and responsible usage

Module 4: Amazon Bedrock API Integration

• Introduction to Amazon Bedrock APIs
  • Understanding model invocation
  • Working with requests and responses
• Integrating foundation models using Python
  • Configuring the AWS SDK for Python (Boto3)
  • Sending prompts and processing responses
• Managing API errors and authentication
• Developing reusable API integration functions

Module 5: Prompt Engineering for Foundation Models

• Understanding prompt engineering principles
  • Writing clear instructions
  • Providing relevant context and examples
• Prompt design techniques
  • Zero-shot and few-shot prompting
  • Role-based and task-specific prompts
• Improving response relevance and consistency
• Testing and refining prompts for business scenarios

Module 6: Text Generation and Content Processing

• Generating text with supported foundation models
  • Content creation and rewriting
  • Summarisation and information extraction
• Developing text-processing applications
  • Document summarisation workflows
  • Question-answering applications
• Configuring supported generation parameters
• Validating model-generated content

Module 7: Conversational AI Development

• Fundamentals of conversational AI
  • Chatbot architecture and interaction flows
  • Managing conversation context
• Building AI-powered assistants
  • Processing user messages
  • Generating contextual responses
• Implementing conversation history
• Handling incomplete or unclear user queries

Module 8: Embeddings and Semantic Search

• Introduction to embeddings
  • Converting text into vector representations
  • Understanding semantic similarity
• Semantic search implementation
  • Preparing and indexing text data
  • Retrieving relevant information
• Exploring vector database integration
• Evaluating search relevance and retrieval quality

Module 9: Retrieval-Augmented Generation (RAG)

• Understanding RAG architecture
  • Document ingestion and preprocessing
  • Text chunking and embedding generation
• Retrieving relevant information
  • Similarity search and context selection
  • Connecting retrieved content to prompts
• Generating context-grounded responses
• Evaluating RAG performance and limitations

Module 10: Amazon Bedrock Knowledge Bases

• Introduction to Amazon Bedrock Knowledge Bases
  • Knowledge base components and architecture
  • Supported data source concepts
• Preparing and connecting organisational documents
  • Data ingestion and synchronisation
  • Retrieval configuration
• Integrating knowledge retrieval with foundation models
• Testing source-grounded question answering

Module 11: Amazon Bedrock Agents

• Introduction to AI agents
  • Agent architecture and capabilities
  • Understanding action groups and tool use
• Configuring agent workflows
  • Defining instructions and supported actions
  • Connecting agents to appropriate services
• Understanding orchestration and task execution
• Applying safeguards and testing agent behaviour

Module 12: Model Customisation and Fine-Tuning Concepts

• Understanding foundation model customisation
  • Prompt-based adaptation
  • Fine-tuning concepts and supported options
• Preparing suitable training datasets
  • Data quality and formatting considerations
  • Data privacy and governance
• Understanding evaluation and model adaptation
• Reviewing service availability and model-specific limitations

Module 13: Security, Governance, and Responsible AI

• AWS identity and access management
  • IAM policies and least-privilege access
  • Secure credential handling
• Data protection and privacy
  • Sensitive information management
  • Secure data processing practices
• Responsible AI principles
  • Bias, harmful outputs, and model limitations
  • Guardrails and output validation
• Understanding monitoring and governance requirements

Module 14: Testing, Evaluation, and Cost Optimisation

• Testing foundation model applications
  • Functional and integration testing
  • Handling invalid requests and service errors
• Evaluating generated responses
  • Relevance, accuracy, and consistency
  • Testing retrieval and prompt quality
• Monitoring performance and latency
• Understanding token usage, pricing, quotas, and cost controls

Module 15: Deployment and Capstone Project

• Designing an AI application architecture
  • Integrating application logic with Amazon Bedrock
  • Configuring secure service access
• Building a foundation model application
  • Developing a chatbot or document assistant
  • Implementing RAG or knowledge retrieval where appropriate
• Testing application functionality and reliability
• Documenting implementation and evaluation results
• Demonstrating the completed project and identifying future improvements