AI-Ready Data

1. Auflage 2026
ISBN: 978-1-394-37105-1
Verlag: John Wiley & Sons Inc

Buch, Englisch, 368 Seiten, Format (B × H): 188 mm x 234 mm, Gewicht: 726 g

Build scalable AI systems by fixing data silos, quality decay, and governance gaps

AI-Ready Data provides a structured roadmap for organizations deploying traditional AI, large language models, and agentic AI systems. Written by Andrew Madson, who has held data strategy leadership roles at Fortune 100 companies including JPMorganChase and MassMutual, the book focuses on the foundational data challenges that undermine AI outcomes. It connects AI engineering, data strategy, and infrastructure planning into a unified approach for building production-grade AI systems.

The book details how to identify and resolve data silos, quality decay, and governance gaps that create hidden costs and erode AI ROI. It covers modern data product architectures and compliance-ready systems designed to accelerate model deployment and reduce technical debt. Each chapter addresses specific operational pain points, from dirty data remediation to building scalable infrastructure that supports traditional ML pipelines, LLM integration, and agentic AI workflows.

Readers will also find:

- Strategies for diagnosing and eliminating data quality decay across enterprise data pipelines before it undermines AI model performance
- Frameworks for building modern data products and architectures that reduce technical debt and accelerate model deployment cycles
- Governance models designed to close compliance gaps and create audit-ready systems for AI initiatives at enterprise scale
- Methods for breaking down organizational data silos that block cross-functional AI adoption in Fortune 100 environments
- Practical approaches to calculating and reducing the hidden costs of dirty data that erode AI return on investment

AI-Ready Data serves business and technology leaders, including CIOs, CDOs, CTOs, and CISOs, as well as data professionals responsible for building and maintaining the data infrastructure behind AI initiatives. It delivers actionable frameworks for resolving data quality, governance, and architecture challenges that directly affect AI system performance.

Autoren/Hrsg.

Madson, Andrew

Fachgebiete

• Mathematik | Informatik EDV | Informatik Daten / Datenbanken Data Mining
• Mathematik | Informatik EDV | Informatik Informatik Künstliche Intelligenz

Inhaltsverzeichnis

Author Bio xviii
Acknowledgments xix
Introduction xx

Chapter 1: Model Access Is Becoming a Commodity, and Data Is a Durable Advantage 1
Chapter 2: The Data-AI Gap 17
Chapter 3: Mapping Your Data for AI 29
Chapter 4: The CRAFT Assessment Framework 45
Chapter 5: CRAFT, Ethics, Privacy, and Compliance 66
Chapter 6: Framework Summary 85
Chapter 7: Raw Data and CRAFT 95
Chapter 8: Systems for Raw Data 112
Chapter 9: Structured Data and CRAFT 126
Chapter 10: Systems for Structured Data 151
Chapter 11: Context Data and CRAFT 168
Chapter 12: Systems for Context Data 186
Chapter 13: Action Data and CRAFT 204
Chapter 14: Systems for Action Data 224
Chapter 15: Assessing and Measuring CRAFT 237
Chapter 16: Organizing for CRAFT 256
Chapter 17: Governance for AI-ready Data 268
Chapter 18: The Economics of CRAFT 283
Chapter 19: Quick Wins and the 90-day Plan 298
Chapter 20: The Last Mile Is Technical and Human 314
Chapter 21: Conclusion: Models Converge. Data Differentiates 324

References 329
Appendix A CRAFT Assessment Matrix 330
Appendix B When Culture Breaks Data Loops 332
Appendix C Entity Integrity 334
Index 336

Über Autor(innen)

ANDREW MADSON is a member of the technical staff at Perplexity AI, where he leads global developer relations. A veteran data strategy leader, Madson has shaped data infrastructure at Fortune 100 organizations including JPMorganChase and MassMutual, as well as high-growth startups like Dremio. He is a regular keynote speaker at industry conferences and reaches hundreds of thousands of data professionals through his blog, podcast, and newsletter.