AI-Ready Data Blueprints. From Raw Data to AI-Driven Innovation Navnit Shukla, Kien Pham, Srikanth Sopirala

Dostępność produktu

Produkt nie został jeszcze oceniony pod kątem ułatwień dostępu lub nie podano żadnych informacji o ułatwieniach dostępu lub są one niewystarczające. Prawdopodobnie Wydawca/Dostawca jeszcze nie umożliwił dokonania walidacji produktu lub nie przekazał odpowiednich informacji na temat jego dostępności.

Spis treści książki

• Foreword
• Preface
  • What Youll Find Inside
  • Who This Book Is For
  • Conventions Used in This Book
  • Using Code Examples
  • OReilly Online Learning
  • How to Contact Us
  • Acknowledgments
• 1. Introduction to AI-Ready Data Foundation
  • Introduction and Market Context
  • What Makes Generative AI Different
    • The Transformer Architecture: The Technical Foundation of GenAI
    • Enterprise Data as the Key Differentiator
    • Contextual Intelligence: A Simple Example
    • Representing Meaning in Vector Space
    • Enterprise Example: Cross-Document Understanding in Customer Support
  • The Evolution of GenAI Applications
    • From Assistants to Agents: The Four Stages of GenAI Evolution
      • Stage 1: AI assistantsinformation retrieval and basic task completion
      • Stage 2: Copilot assistantscollaborative work with humans
      • Stage 3: RAG-based agentsenhanced knowledge and contextual understanding
      • Stage 4: Agentic AIautonomous decision making and action
    • The Increasing Complexity of Data Requirements
  • What Leading Organizations Are Building Today
    • The Production Reality Check
    • Emerging GenAI Architectural Patterns
      • Pattern 1: Knowledge graphs for contextual intelligence
      • Pattern 2: Event-driven foundations for real-time AI
      • Pattern 3: Lakehouse as unified AI platform
      • Pattern 4: Semantic search as the new query layer
      • Pattern 5: Agent-ready digital properties
    • From Patterns to Practice
  • Development Cycles: ML Versus GenAI
    • Traditional Machine Learning Development Cycle
    • GenAI Development Cycle
    • Data Infrastructure Implications
  • Architecture Evolution in Practice: From Traditional ETL to GenAI-Ready Pipelines
    • Key Differences Between Traditional and GenAI Data Foundations
    • Real-World Example: Evolving from Kimball to Medallion to GenAI-Ready Architecture
      • The challenge: multi-billion-row fact tables in the GenAI era
      • The solution: a hybrid approach for GenAI readiness
      • GenAI-enhanced data pipeline architecture
      • Results and benefits
      • Scaling strategies for the AI era
      • Why this matters: The data foundation for AI success
  • Preparing for the Agent-Driven Future
    • Jeeves Does the Shopping: The Agent-Mediated Consumer Experience
    • Search Engine Optimization: From Keywords to Agent Optimization
    • Agents Have No Allegiance: Preparing for Radical Transparency
    • Business Autopilot: Autonomous Operations and Decision Making
    • Data Readiness Checklist for the Agent-Driven Future
  • Blueprint: Implementation Guidance
    • Where to Start
      • If youre just starting (experimentation stage)
      • If you have basic GenAI running (pilot stage)
      • If youre scaling production (growth stage)
    • The Decision Tree: Which Pattern First?
    • Your Action Plan
  • Summary
• 2. Data Framework for GenAI and Agentic AI Applications
  • Introduction: Building the Foundation for AI-Ready Data
  • The Evolution of Data Frameworks
    • Early Days (20152019)
    • Transition Period (20202022)
    • GenAI Era (2022Present)
    • Agentic AI Era (2024Present)
  • The Need for a New Approach: Core Requirements for AI-Ready Data
    • Capturing Business Logic and Context
    • Ensuring Data Quality and Consistency
    • Managing Complexity and Diversity
    • Maintaining Security, Compliance, and Privacy
    • Enabling Information Sharing and Collaboration
    • Supporting Scale and Performance
    • Managing Data as a Strategic Product
    • Empowering Users with Documentation and Guidance
  • A Core Framework for AI-Ready Data
    • Capturing Business Logic and Context
      • Key elements
      • Implementation best practices
      • Real-world examples
      • Summary
    • Ensuring Data Quality and Consistency
      • Key elements
      • Implementation best practices
      • Real-world examples
      • Summary
    • Managing Complexity and Diversity
      • Key elements
      • Implementation best practices
      • Real-world examples
      • Summary
    • Maintaining Security, Compliance, and Privacy
      • Key elements
      • Implementation best practices
      • Real-world examples
      • Summary
    • Enabling Information Sharing and Collaboration
      • Key elements
      • Implementation best practices
      • Real-world examples
      • Summary
    • Supporting Scale and Performance
      • Key elements
      • Implementation best practices
      • Real-world examples
      • Summary
    • Managing Data as a Strategic Product
      • Key elements
      • Implementation best practices
      • Real-world examples
      • Summary
    • Empowering Users with Documentation and Guidance
      • Key elements
      • Implementation best practices
      • Real-world examples
      • Summary
  • AI-Ready Data Blueprints for the Data Framework: Practical Implementation Guide
    • Blueprint 1: Business Context Intelligence Engine
      • Phase 1: Context foundation (months 12)
      • Phase 2: Intelligent context injection (months 34)
    • Blueprint 2: Adaptive Data Quality Orchestration
      • Phase 1: Quality assessment and monitoring (months 12)
      • Phase 2: Self-healing data systems (months 34)
    • Blueprint 3: Orchestrating Data Diversity and Complexity
      • Phase 1: Data classification, integration, and cataloging (months 12)
      • Phase 2: Diversity, bias auditing, and dependency management (months 34)
    • Blueprint 4: Security-First AI Data Platform
      • Phase 1: Security architecture and access controls (months 12)
      • Phase 2: Compliance automation and privacy preservation (months 34)
  • Summary
• 3. Data Wrangling and Data Preparation for GenAI and Agentic AI Applications
  • The Enterprise Challenge
    • Purpose and Audience
    • The Market Imperative
    • The Strategic Advantage
    • Transformation Journey Preview
  • Understanding the Paradigm Shift
    • The Established Machine Learning Data Pipeline
    • The Tabular Data Paradigm and Its Constraints
    • The Conceptual Revolution in Data Processing
    • New Processing Paradigms and Technical Requirements
      • Document parsing and semantic chunking
      • Entity extraction and vector embedding
      • Multimodal integration and semantic normalization
    • The Business Impact of the Paradigm Shift
    • Self-Assessment: Where Is Your Organization Today?
  • Building Blocks of GenAI-Ready Data
    • Semantic Understanding Fundamentals
      • From syntax to semantics
      • Ontologies and knowledge modeling
      • Vector embeddings and similarity search
      • Embedding generation and management
      • Vector search and retrieval
    • Knowledge Graphs and Relationship Modeling
      • Knowledge graph fundamentals
      • Implementation approaches
    • Real-Time Data Processing Requirements
      • Streaming semantic processing
      • Consistency and currency management
    • GenAI Data Preparation Maturity Model
    • Case Study: Retail Organization Transformation
      • Initial state
      • Transformation approach
      • Outcomes and lessons
  • The Semantic Layer Architecture
    • Architectural Overview and Core Principles
    • Data Foundation Layer: Building on Lakehouse Architecture
      • Key components
      • Implementation considerations
    • Metadata and Ontology Management: Creating Semantic Understanding
      • Key components
      • Implementation considerations
    • Transformation and Enrichment Pipeline: Adding Intelligence to Data
      • Key components
      • Implementation considerations
    • Vectorization and Indexing Infrastructure: Enabling Semantic Search
      • Key components
      • Implementation considerations
    • APIs and Reasoning Layer: Enabling AI Consumption
      • Key components
      • Implementation considerations
    • Component Interactions and Data Flows
      • Ingestion and enrichment flow
      • Query and retrieval flow
      • Knowledge evolution flow
    • Implementation Considerations and Common Patterns
      • Cloud native implementation
      • Hybrid integration pattern
      • Specialized domain pattern
      • Evolutionary implementation
  • AWS Reference Architecture and Implementation
    • AWS Ecosystem Overview for GenAI Data Preparation
    • Key AWS Services and Capabilities
    • Integration Patterns and Service Selection
      • RAG implementation pattern
      • Unified data and AI pattern
      • Multimodal content processing pattern
  • Governance, Quality, and Observability
    • Expanded Quality Dimensions for GenAI Data
      • Traditional quality dimensions
      • GenAI-specific quality dimensions
      • Quality assessment approaches
    • Governance Framework Essentials
      • Data and knowledge ownership
      • Policy management
      • Process governance
      • Technology governance
      • Measurement and reporting
    • A Governance and Quality Monitoring Framework
      • Policy layer
      • Process layer
      • Technology layer
      • Measurement layer
      • Improvement layer
      • Integration approach
    • Monitoring and Observability Approaches
    • Regulatory Compliance and Assurance
    • Case Study: Financial Services Governance Implementation
      • Initial governance challenges
      • Governance solution
      • Outcomes and lessons
  • Advanced Topics: Agentic AI Data Requirements
    • Agentic AI Fundamentals and Data Implications
      • Continuous learning, real-time ingestion, and embedding freshness
      • Agent-specific retrieval and context management
    • Enterprise Implementation Examples
  • The Future: Autonomous and AI-Assisted Data Wrangling
    • AI-Powered Data Discovery and Intelligent Classification
    • Conversational Data Preparation and Natural Language Interfaces
    • Autonomous Quality Assurance and Self-Healing Systems
    • Future Directions and Research Frontiers
  • Action Plan
    • Key Principles
    • Immediate Next Steps by Maturity Level
    • Long-Term Strategic Considerations
  • Summary
• 4. Data Governance, Security, Compliance, and Orchestration for GenAI
  • Data Governance and Data Security for AI Applications
    • The AI Data Governance Operating Model
    • Differences in Governance of Structured and Unstructured Data
  • Data Stewardship and Metadata Management
    • Making Data Accessible for Humans and Agentic AI
    • Testing Our Tools
  • Data Quality
  • Responsible AI
    • Fairness
      • Tools to detect bias
      • How to handle bias
    • Transparency
      • Tracing with LLMs
      • Capturing and exporting telemetry
    • Accountability
      • Responsibility
      • Correction
    • Privacy and Security
      • Encryption at rest and in transit
      • Sensitive data identification and redaction
      • Access control in vector stores
    • Reliability
    • Human Oversight
  • Data Privacy and Security for GenAI Applications
    • Sensitive Information Protection
    • Topic Restriction and Word Filtering
    • Filtering Tools
    • End-to-End Data Protection
  • LLMOps: AI Workflow Orchestration
    • Orchestration Patterns for AI Applications
      • Linear (chained) orchestration
      • Modular and concurrent orchestration
    • Agentic Patterns
      • Supervisor agent with specialized agents
      • Example use case and implementation: ecommerce customer support
  • Summary
• 5. Knowledge Bases and Vector Databases
  • The GenAI Data Challenge
    • Knowledge Bases: Data Organization and Storage
    • Vector Databases: Data Representation and Similarity Search
    • Retrieval-Augmented Generation: Data Retrieval and Context Assembly
    • Why These Technologies Are Essential
  • Knowledge Base Fundamentals: From Data to Knowledge
    • Data Architecture for Knowledge Bases
      • Knowledge base components through a data lens
        • Data foundation
        • Metadata and ontology management
        • Transformation and enrichment pipeline
        • Vectorization and indexing infrastructure
        • APIs and reasoning
      • Governance: Embedded throughout
      • Evolution from traditional to GenAI knowledge bases
      • Real-world example
    • Data Preparation Pipeline for Knowledge Bases
      • Seven-stage data transformation
      • Real-world example
    • Data Types and Management Strategies
      • Unstructured data: The 8090% challenge
        • Data quality challenges by type
        • Data handling strategies
      • Structured data: Precision and relationships
      • Hybrid data strategies
      • Real-world example
    • Data Quality: The Make-or-Break Factor
      • Common data quality problems in knowledge bases
      • Data quality metrics for continuous monitoring
  • Vector Database Fundamentals: Data Representation
    • Embeddings and Data Quality
    • Vector Database Infrastructure
    • Open Source Solutions
    • Data Indexing Algorithm Comparison
    • Data Compression Techniques
    • Data Governance Feature Comparison
    • Selecting Based on Data Requirements
  • RAG: The Data Retrieval Layer
    • Why RAG? The Data Access Problem
    • The RAG Architecture: Data Flow
      • Data-centric benefits of RAG
      • RAG architecture evolution: From simple to agentic
    • Data Chunking: Critical for RAG Performance
      • Comprehensive data chunking strategies
      • Data-specific chunking guidelines
      • Chunk metadata enrichment
      • Overlap strategy for data continuity
      • Data retrieval optimization for RAG
      • Data reranking strategies
      • Query transformation for better data retrieval
      • Prompt engineering for RAG: optimizing context delivery to the LLM
    • Data Efficiency and Cost Optimization
      • Data caching strategies
      • Data retrieval cost management
      • Real-world example
      • Data retrieval optimization techniques
      • Data caching strategies
      • Data partitioning for scale
      • Data quality monitoring in production
      • Data drift detection
    • End-to-End Data Flow: KB Vector DB RAG
      • Advanced multimodal integration
      • Real-time data synchronization
      • Data strategy best practices
  • Summary
• 6. AI Application Optimization for Production Readiness
  • The Journey from Prototype to Production
    • Preparing Your AI Application for Production Success
    • The Twin Pillars of AI Optimization
  • Engineering Efficient Data Workflows for Production AI Systems
    • Data Quality Assessment and Preprocessing Fundamentals
      • Business outcomes drive data quality requirements
      • Efficient data management through modular design
      • Best practices for modular data design
    • Understanding Context Windows and Their Implications
    • Handling Large Input Contexts
    • Optimizing Inference Quality Using Automated Reasoning
      • Token-based reasoning
      • Architectural reasoning
        • Chain-of-thought
        • Tree-of-thought
        • Graph-of-thought
        • Reasoning framework selection
      • Processing patterns
  • Metadata Readiness for Agentic AI Systems
    • The Challenges of Scale and Complexity
      • The complexity within
      • Why traditional approaches fall short
      • A new paradigm: Intelligent semantic metadata
    • Pattern: Leveraging Raw Files Directly
      • Why organizations choose this path
      • The emergence of intelligent semantic metadata
    • Building Data-Aware Agents Through Intelligent Semantic Metadata
      • Core capabilities
      • Operational efficiency gains
      • The value proposition
    • Example Implementation: Leveraging Raw Files Directly
      • AI agent steps
      • The intelligence in the approach
    • From Raw Content to Semantic Intelligence: Constructing the Metadata Layer
      • Start with the end in mind
      • Designing the graph schema
      • Embracing iteration over perfection
      • Extracting and generating metadata
      • Leveraging prompt engineering
      • Handling large files
      • Loading and activating the metadata layer
      • The iterative lifecycle
  • Key Capabilities for Production Readiness for Agentic AI Platforms
  • Leading AI Agent Platform Choices
    • Core Runtime and Deployment Strategy
    • Framework, Tool Integration, and Protocol Support
    • Memory and Knowledge Management Capabilities
    • Security and Compliance Features
    • Observability and Quality Assurance Support
  • Summary
• Index

pokaż cały spis treści

ukryj recenzje