Notes from a practicing enterprise architect. Reading log, mostly books. Slow-moving, occasionally grumpy, no promises about publishing schedule.
The best enterprise architecture books in 2026
A practitioner's ranked, opinionated buying guide to the enterprise architecture books worth your money in 2026 — the classics that still hold up, the TOGAF certification track, and the first books that take agentic AI seriously. Each one links to a full review.
Start here
TOGAF 10 self-study path: from OGEA-101 to OGEA-102
A practical, budgeted study plan for the OGEA-101 (Foundation) and OGEA-102 (Practitioner) exams. The books, the Series Guides, the practice questions, and what to ignore.
Enterprise Architecture As Strategy: the book that named the Operating Model
Ross, Weill, and Robertson wrote the only EA book that ever got serious boardroom traction. Twenty years on, the Four Operating Models are still the most useful diagnostic I have.
Agentic Architectural Patterns: the first serious patterns catalogue for agent systems
Arsanjani and Bustos have written the first patterns catalogue that takes agentic systems seriously as an architectural problem. It is dense, it is opinionated, and it is the book I have been waiting for.
All notes
Is TOGAF certification worth it in 2026? An honest take
TOGAF certification is a market filter, not a quality signal. Whether it is worth your money and study time depends entirely on which segment of the market you are targeting — and the answer differs more than the prep-course marketing suggests.
Enterprise architect vs solution architect: the honest difference
The org chart says one reports to the other. That's the least interesting distinction. The honest difference is altitude, what each is accountable for when something breaks, and which reading builds the muscle each role actually needs.
When the review board becomes the bottleneck: architecture governance at agent speed
Architecture governance was built for human-speed change: ADRs, review boards, approval gates. Agentic delivery breaks every assumption underneath it. The gate is now the bottleneck, and the fix is not a faster gate.
TOGAF in the AI era: what still applies, what doesn't
A practitioner's honest audit of which parts of TOGAF survive agentic AI and which strain — the ADM loop and governance hold, big-design-up-front does not — and where you have to supplement the framework rather than defend it.
How to become an enterprise architect: a 2026 reading path
The honest route from senior engineer into the EA role — what the job actually is, the skills that decide who lands the title, and a sequenced reading path that treats TOGAF as signalling, not salvation.
The best enterprise architecture books in 2026
A practitioner's ranked, opinionated buying guide to the enterprise architecture books worth your money in 2026 — the classics that still hold up, the TOGAF certification track, and the first books that take agentic AI seriously. Each one links to a full review.
Classic vs bleeding edge: how EA reading has evolved
The EA canon split, quietly, around 2023. There is the pre-agentic literature and the post-agentic literature, and the most interesting books are the ones that bridge them. A short tour.
TOGAF 10 self-study path: from OGEA-101 to OGEA-102
A practical, budgeted study plan for the OGEA-101 (Foundation) and OGEA-102 (Practitioner) exams. The books, the Series Guides, the practice questions, and what to ignore.
Agentic Architectural Patterns: the first serious patterns catalogue for agent systems
Arsanjani and Bustos have written the first patterns catalogue that takes agentic systems seriously as an architectural problem. It is dense, it is opinionated, and it is the book I have been waiting for.
Enterprise Architecture in the Age of Agentic AI: an honest first reading
Maher Dahdour's 2026 book is the first serious attempt at an EA framework for a world where agents are first-class citizens. It is uneven. It is also necessary.
Mastering TOGAF 10: the study companion the standard needed
Roshan Gavandi's book is the closest thing TOGAF 10 has to a serious self-study companion. Here is what it does well, where it falls short, and how I would use it alongside the standard itself.
Building Evolutionary Architectures, 2nd ed: fitness functions finally get the treatment they deserve
Ford, Parsons, and Kua's second edition is a serious upgrade. The fitness-function concept gets the room it needs, and the book is now the practical reference I wanted the first edition to be.
Team Topologies: the org-design vocabulary the industry needed
Skelton and Pais's small book has done more for honest conversations about engineering org design than any framework in the last decade. Here is what holds up, what does not, and how I have actually used it.
The Software Architect Elevator: the EA book I quote most in conversation
Gregor Hohpe wrote the book most practising architects wish their CIO had read. Five years on, the elevator metaphor still does most of the heavy lifting in conversations about what EA is actually for.
Designing Data-Intensive Applications: the only systems book I re-read on purpose
Kleppmann's book has become the de facto common vocabulary for distributed systems. Eight years on, it is still the one I lend out most, and the only one I have re-read on purpose three times.
The TOGAF Standard, 10th Edition: a review from someone who actually reads it
Most people who recommend TOGAF have skimmed it. Most who criticize it have skimmed it. I read the 10th Edition cover to cover so you can decide whether you need to.
What I read in 2025: an EA's top 10
Ten books that earned a place on my desk in 2025, with short blurbs and links to the longer reviews. Half are classics I went back to. Half are too new to be classics yet.
Enterprise Architecture As Strategy: the book that named the Operating Model
Ross, Weill, and Robertson wrote the only EA book that ever got serious boardroom traction. Twenty years on, the Four Operating Models are still the most useful diagnostic I have.
The Phoenix Project, ten years later: still the cheapest org-design lesson around
Re-reading Gene Kim's parable now that 'DevOps' is a job title, a vendor category, and an awkward team name. What still lands, what feels dated, and why I keep handing it to new engineering managers.
