MAINTAINED · v2026.09.1 · CURRENT AS OF 2026-09-11
The AI Transformation Field Guide
From AI pilots to operational transformation.
A maintained, ungated reference for leaders moving state and local government, public authorities, and the companies that serve them from AI experiments into governed, measurable production. It covers the whole middle of the work that pilots skip: operating models, agents and their controls, data and architecture, security and identity, governance, economics and measurement.
Written for executives and senior operators who have to decide what reaches production, what it may touch, and how they will know it worked. Every topic is written at executive length, ten to fifteen minutes each, and cites primary sources rather than vendor material.
THREE WAYS TO USE IT
Read it, download it, or hand it to your model
CONTENTS
Explore the guide
Every card is one section. Read them in order, or go straight to the domain you are deciding about this quarter. Catalog sections are numbered in reading order.
Start Here
Read this first. It explains what the guide is for and how to use it.
1 sectionThe Topic Catalog
Twenty-one domains, numbered in reading order. Each one is written at executive length.
21 sections- The Topic Catalog · 1Enterprise AI Strategy & Operating ModelCentralized, federated or hub-and-spoke. Portfolio management, build versus buy, multi-model strategy, and how transformation itself is governed.
- The Topic Catalog · 2Agentic AIWhat an agent actually is, how agent architectures differ, and how much authority a workflow can safely hand to one.
- The Topic Catalog · 3MCP & Agent InteroperabilityModel Context Protocol, the security and governance questions it raises, and the wider agent-to-agent interoperability landscape.
- The Topic Catalog · 4AI-Native Software EngineeringFrom autocomplete to autonomous coding agents: specification-driven development, AI code review, generated tests, and what changes in the lifecycle.
- The Topic Catalog · 5Future Engineering Workforce & Organizational DesignWhich engineering skills appreciate and which commoditize, the junior pipeline problem, and how team topology and hiring change.
- The Topic Catalog · 6AI for Professional & Knowledge WorkersThe maturity ladder from chat to autonomous business processes: enterprise assistants, deep research, meeting intelligence, departmental digital workers.
- The Topic Catalog · 7Workflow AutomationRPA, BPM, API automation and agentic workflows compared, and where deterministic automation still wins.
- The Topic Catalog · 8Enterprise Knowledge & Data ArchitectureRetrieval-augmented generation, embeddings, knowledge graphs, permissions-aware retrieval, data classification and memory architecture.
- The Topic Catalog · 9Foundation Models & AI EngineeringFrontier and open-weight models, context windows and tokens, routing and fallback, fine-tuning, distillation and small models at the edge.
- The Topic Catalog · 10Prompt Engineering vs. Context EngineeringPrompting fundamentals, and why the architecture of context around the model matters more than the prompt itself.
- The Topic Catalog · 11Enterprise AI Gateway & Control PlaneAI gateways as the enterprise control point: centralized policy, rate limits, data loss prevention, cost controls and architecture choices.
- The Topic Catalog · 12AI Evaluation & Quality EngineeringEvals and golden datasets, trajectory evaluation for agents, groundedness and LLM-as-judge, red teaming and continuous production evaluation.
- The Topic Catalog · 13AI SecurityThe OWASP Top 10 for LLM applications, prompt injection, excessive agency, poisoning, supply-chain risk and zero-trust architecture for agents.
- The Topic Catalog · 14Agent Identity & AuthorizationNon-human identity for agents, delegated authorization, just-in-time access, least privilege, transaction limits and audit attribution.
- The Topic Catalog · 15AI Governance, Legal, Privacy & ComplianceNIST AI RMF, ISO/IEC 42001, the EU AI Act, intellectual property risk, system inventories and risk classification.
- The Topic Catalog · 16AI FinOps & EconomicsToken economics and cost per outcome, routing and caching as cost controls, chargeback, showback and runaway agent cost risk.
- The Topic Catalog · 17AI Observability & AgentOpsTracing agent trajectories, LLMOps and AgentOps, incident response and audit replay for agent actions.
- The Topic Catalog · 18AI-Native Product DesignBeyond the chatbot: progressive autonomy and approval UX, citations, provenance, confidence indicators and designing for AI failure.
- The Topic Catalog · 19Organizational Adoption & Change ManagementRole-based AI literacy, champions and communities of practice, displacement fear, change management and acceptable use.
- The Topic Catalog · 20AI Measurement & Business ValueDORA and AI-specific engineering metrics, knowledge-worker productivity, enterprise value metrics, and why seat counts are vanity metrics.
- The Topic Catalog · 21Emerging Technology RadarWhat is near-term, developing or speculative, why each item matters, and the concrete signal that would show it is becoming real.
Putting It Into Practice
Instruments for turning the catalog into a plan, a ranked reading list and a score.
4 sections- Putting It Into PracticeThe 90-Day Executive AI Learning PlanThe topic catalog sequenced into a working study schedule at three to five hours a week, with one experiment and one hard question each week.
- Putting It Into PracticeThe 25 Concepts I Would Learn FirstIf time only permits twenty-five items from the whole document, this is the ranked list, cutting deliberately across domains.
- Putting It Into PracticeExecutive Technology RadarOne decision instrument: current importance, two to three year potential, and a recommended action for every topic and radar item.
- Putting It Into PracticeThe Enterprise AI Capability ModelA scoring instrument across fifteen capability areas and four levels. Score honestly, then see which gaps are holding the others back.
Reference
The sources behind the catalog and the vocabulary it uses.
2 sections