AI-NATIVE BUSINESS INTELLIGENCE Volume 1 · Foundations, Architecture, and Transformation: from classical data warehousing to agentic, AI-first analytics — how get there the system

Your dashboards show you the number. Who explains the why?Every BI system built in the last twenty years was designed to answer "what happened." The AI era demands more: systems that investigate causes on their own, explain their answers, and act - without losing the discipline that made classical BI trustworthy. This book builds that bridge, end to end.AI-Native Business Intelligence, Volume 1 is a complete, practice-first course in three parts: PART I - Classical BI, rebuilt properly. Inmon vs. Kimball, dimensional modeling, slowly changing dimensions, ETL/ELT, data quality, MDM, and governance - the foundation your future agents will stand on, taught through one continuous retail case.PART II - The AI-Native architecture. The six-layer reference model with the semantic layer at its heart: metrics as code, MCP as the controlled door, conversational and agentic BI, RAG and knowledge graphs, hallucination control, compile-time security, and FinOps for agent workloads. You'll learn why "wrap an LLM around the database and hope" fails - and what to design instead.PART III - Transformation in practice. Greenfield, brownfield, and hybrid playbooks with roadmaps, risk registers, and rollback criteria - capped by a full multinational, multi-currency capstone (Truong Son Group) you can rebuild for your own organization, exercises and reference solutions included.Inside this volume- 35 chapters, 60+ original architecture figures, and chapter quizzes with full answer keys- Ready-to-adapt templates: semantic-layer YAML, MCP server configuration, eval suites with "trap" questions, Bus Matrix, design checklists- A 90-day / 6-month / 12-month migration playbook with quantitative stop-and-rollback criteria- 30/60/90-day self-study paths for newcomers, BI veterans, and decision-makersWritten for data engineers, architects, analytics leaders, and CTOs who refuse to choose between AI capability and engineering discipline. No prior AI experience required; every concept is built from the ground up.Volume 1 is the thinking book - what to build and why. Volume 2 (Design & Implementation) is the building book. They pair, but each stands alone.