AI Architecture

AI Engineering Books: The 3 That Make an AI Architect (and What None of Them Teach)

Software architects and AI engineers are solving the same problems with different vocabularies, and most reading lists make it worse by putting them on separate shelves. I mapped three books against each other to see what really overlaps. Four ideas sit in all three, eleven more are shared by two, and six things an AI architect needs are in none of them.

Gurram Poorna Prudhvi

Lead AI Engineer

Intermediate
Nov 16, 2026
11 min read
THE AI ARCHITECT'S READING MAPthree books · what overlaps · what none of them teachFundamentalsRICHARDS & FORD · 2EThe Hard PartsFORD · RICHARDS · SADALAGE · DEHGHANIAI EngineeringCHIP HUYENArchitecture characteristicsCohesion · coupling metricsNine architecture stylesChoosing a styleRisk storming · leadershipGranularity driversDecomposing monolithsPulling apart dataEight saga patternsData meshFoundation models · evalsPrompting · RAG · agentsFinetuning · datasetsInference optimizationGuardrails · routers · feedbackARCH BOOKSADRsStamp couplingArchitecture quantumObservabilityLatency · cost -ilitiesLayered AI stackEvolve in small stepsDataset ownershipExact · semantic cacheShared model gatewayData flywheelALL THREEEvery choice a trade-offFitness functions ≈ evalsOrchestrated workflowsSchemas as contractsBEYOND ALL THREEMCP · agent protocolsDurable, long-running agentsAI governance · EU AI ActGPU capacity · FinOpsOn-call for probabilistic systemsAI product senseaiengineerinsights.com · mapping is ours, from the three books' tables of contents, Nov 2026

TL;DR: The three AI engineering books worth reading together are Fundamentals of Software Architecture (how to think), Software Architecture: The Hard Parts (what breaks when systems are distributed) and Chip Huyen's AI Engineering (what changes when a component is a model). All three share four ideas: every choice is a trade-off, fitness functions are evals, workflows need an orchestrator, and schemas are contracts. None of them covers agent protocols, long-running agents, governance, GPU budgets, on-call for systems that don't repeat themselves, or product sense. Those you learn from docs and practice.

Explore this article with

1. Why These Three Books

Most lists of AI engineering books are ten titles long and tell you to read all of them. This one is three, chosen because each answers a different question, and because together they cover the job people now call AI architect: designing systems where some of the components are models.

BookAuthorsPublishedThe question it answers
Fundamentals of Software Architecture, 2nd editionMark Richards, Neal FordMarch 2025How to think like an architect: characteristics, modularity, styles, decisions
Software Architecture: The Hard PartsNeal Ford, Mark Richards, Pramod Sadalage, Zhamak DehghaniOctober 2021What breaks when a system is distributed: coupling, data, transactions, contracts
AI EngineeringChip HuyenDecember 2024What changes when a component is a foundation model: evals, RAG, agents, inference

The first two are from the same family of authors and were written to be read in that order. The third comes from a different community. That gap is the interesting part: where they agree, they usually don't know they agree.

Links to the publisher's pages: Fundamentals of Software Architecture, Software Architecture: The Hard Parts, AI Engineering. Chapter names below are from their published tables of contents.

2. The Idea in the Middle: An Eval Is a Fitness Function

Start with the overlap that matters most. In the architecture books, a fitness function is an automated check that tells you whether a system still has a property you care about. Chapter 6 of Fundamentals is called "Measuring and Governing Architecture Characteristics", and The Hard Parts attaches a fitness function to every decomposition pattern it describes.

In AI Engineering, an eval is an automated check that tells you whether a model-backed feature still does what you need. Huyen spends two chapters on it, covering exact evaluation, AI as a judge, and how to design an evaluation pipeline.

One community writes these checks against coupling and latency. The other writes them against faithfulness and tone. They are the same tool, used for the same reason: you can't keep a property you don't test for. An architect who already runs fitness functions in CI knows how to run evals in CI. An AI engineer who already gates releases on an eval set is already doing architecture governance, and can extend the habit to cost, latency and coupling.

Three more ideas sit in the centre with it.

Shared ideaIn FundamentalsIn The Hard PartsIn AI Engineering
Every choice is a trade-offStated as a law of software architecture, and the reason for the phrase "least worst architecture"The closing chapter is called "Build Your Own Trade-Off Analysis"Build versus buy, RAG versus finetuning, cost versus latency versus quality
Fitness functions ≈ evals"Governance and Fitness Functions" for measuring architecture characteristicsFitness functions attached to each decomposition patternTwo chapters on evaluation, including AI as a judge and designing an evaluation pipeline
Orchestrated workflowsMediated event-driven architecture and orchestration-driven stylesOrchestration versus choreography, and eight transactional saga patternsAgent planning and "AI pipeline orchestration"
Schemas as contractsContracts between a microkernel core and its plug-insStrict versus loose contracts, and stamp couplingStructured outputs from a model

The workflow row deserves a second look. A saga is a long transaction split across services, with a plan for what to do when a step fails. An agent run is a long task split across model and tool calls, with the same need. The eight saga patterns are a ready-made vocabulary for multi-agent orchestration: who coordinates, whether steps wait for each other, and what "done" means when something fails halfway.

3. What Two Books Share: The Ideas That Transfer

These are the overlaps between pairs. Each is an architecture idea with a direct use in an AI system, or the reverse.

Shared byIdeaWhat it means for an AI system
Both architecture booksADRsA written record of a decision and why. Use them for model choice, RAG versus finetuning, and vendor lock-in
Both architecture booksArchitecture quantum; static and dynamic couplingThe smallest part you can deploy on its own. A feature and the model it calls are coupled at runtime, so they fail together
Both architecture booksStamp couplingPassing a whole record when one field is needed. With an LLM this is paid for per token
Fundamentals + AI EngineeringObservabilityA request is a tree of model and tool calls, so it has to be traced, not just timed
Fundamentals + AI EngineeringLatency and cost as characteristicsTreat them like any other "-ility": name them, measure them, pick a few to protect
Fundamentals + AI EngineeringLayered thinkingHuyen's three layers of the AI stack are a layered architecture, with the same rules about what may call what
Fundamentals + AI EngineeringEvolve in small stepsHer architecture chapter adds one piece at a time: context, guardrails, router and gateway, cache, agent patterns
The Hard Parts + AI EngineeringData ownershipWho owns a table becomes who owns a dataset, an index, or an eval set
The Hard Parts + AI EngineeringCachingReplicated caching between services; exact and semantic caching in front of a model
The Hard Parts + AI EngineeringShared service or gatewayThe reuse patterns (shared library, shared service, sidecar) are the options for a model gateway
The Hard Parts + AI EngineeringAnalytical data and the data flywheelData mesh on one side, user feedback feeding the next model on the other

The stamp coupling row is my favourite example of an old idea with a new price. Sending a whole customer record to a service that needs one field has always been wasteful. Sending a whole document to a model that needs one paragraph is the same mistake, except now every extra token is billed and makes the answer worse. See the fallacies of LLM systems for the evidence on long context.

For the observability row, AI agent observability shows what a trace of one agent request looks like. For the gateway row, Conway's law for AI covers who should own it.

4. What Each Book Gives You Alone

BookOnly hereRead it first if
FundamentalsIdentifying and limiting architecture characteristics; cohesion, coupling and connascence; the catalog of styles from layered to microservices; diagramming, risk, negotiation and leadershipYou have never had to justify an architecture to other people
The Hard PartsGranularity disintegrators and integrators; decomposing a monolith and its database; data ownership; distributed transactions; the eight saga patterns; data meshYou are splitting a system up, or debugging one that was split badly
AI EngineeringHow foundation models work and are sampled; evaluation methodology; prompt engineering and prompt attacks; RAG and agents; finetuning; dataset engineering; inference optimizationYou are putting a model into a product for the first time

Two notes on AI Engineering. It is not a book about one vendor's API, so it has aged better than most writing on the subject. And its last chapter, "AI Engineering Architecture and User Feedback", is the one closest to the other two books: it builds a reference architecture one step at a time, which is how an architect would want it explained.

Get the AI engineering newsletter

Practical, engineering-first breakdowns on AI agents, LLMs, and breaking into the field — plus the free roadmap PDF when you join. No spam.

By subscribing you agree to receive emails from AI Engineer Insights. Unsubscribe anytime. See our Privacy Policy.

5. What None of Them Teach

Six things an AI architect deals with are outside all three books. That is not a criticism: books take years, and some of this didn't exist when they were written. You learn these from documentation and from running systems.

TopicWhy the books don't cover itStart here
MCP and agent protocolsThe books predate most of the toolingWhat is MCP?
Durable, long-running agentsSagas cover long transactions, not agents that run for hours and must resumeAgent runtimes explained
AI governance and the EU AI ActNone of the three is a compliance bookAI guardrails
GPU capacity and FinOpsInference optimization is covered; capacity planning and budgets are notLLM routing
On-call for probabilistic systemsRunbooks assume the same input gives the same outputThe fallacies of LLM systems
AI product senseKnowing which problems a model should not be used forConway's law for AI

6. Reading Order: Two 12-Week Plans

Which book you start with depends on which half you already have. Both plans end the same way: build something and write the decisions down.

If you are a software architect moving into AI

You have read the two architecture books, or lived them. Start with the model.

WeeksReadWhy
1–2AI Engineering ch. 1–2What a foundation model is and why its output varies
3–4AI Engineering ch. 3–4Evaluation. Read these as a book about fitness functions
5AI Engineering ch. 5Prompting, including prompt injection
6–7AI Engineering ch. 6RAG and agents
8AI Engineering ch. 10The reference architecture: this is the chapter closest to your day job
9–10AI Engineering ch. 7–9, skimmedFinetuning, datasets and inference: know what exists and what it costs
11The Hard Parts ch. 11–13, rereadWorkflows, sagas and contracts, this time with agents in mind
12BuildOne AI feature with an eval suite in CI and an ADR for each big choice

If you are an AI engineer growing into architecture

You can ship a RAG feature. What you are missing is the language for defending a design and the patterns for when it has to be split across teams.

WeeksReadWhy
1–2Fundamentals ch. 1–5Architectural thinking, modularity, and how to pick characteristics
3Fundamentals ch. 6–8Fitness functions, scope, components. You already write evals; this is the same habit
4–5Fundamentals, the styles chaptersRead event-driven and microservices closely; skim the rest
6Fundamentals, the chapters on decisions and riskADRs and risk storming
7–8The Hard Parts ch. 1–4 and 7Coupling, quanta and service granularity
9The Hard Parts ch. 8–10Reuse patterns, data ownership, data access
10The Hard Parts ch. 11–13Orchestration versus choreography, sagas, contracts
11The Hard Parts ch. 14–15Analytical data and trade-off analysis
12ApplyWrite ADRs and fitness functions for an AI system you already run

If you are earlier than either, start with the AI engineer roadmap and come back to this list at phase 3.

Go deeper on AI engineering books and the AI architect reading map with AI

Ask your next question with this article as context. Opens in a new tab; you can edit the question before sending.

  • Which of these three books should I read first?
  • Turn our evals into architecture fitness functions
  • Write an ADR for a model or RAG decision
  • Map saga patterns onto our agent workflow

7. Frequently Asked Questions

What is the best AI engineering book?

For building products on foundation models, AI Engineering by Chip Huyen (O'Reilly, December 2024) is the most complete single book. It covers evaluation, prompt engineering, RAG and agents, finetuning, dataset engineering, inference optimization and a reference architecture. Pair it with a software architecture book if you design systems and not just features.

Do AI engineers need to learn software architecture?

Yes, once a feature has to run in production alongside other systems. Trade-off analysis, coupling, data ownership, contracts and distributed workflows all apply to AI systems, and the architecture books explain them more thoroughly than AI books do. Fundamentals of Software Architecture is the place to start.

What is an AI architect?

An AI architect designs systems in which some components are AI models. The role combines classic software architecture (characteristics, coupling, data, trade-offs) with AI engineering (evaluation, retrieval, agents, inference cost). There is no standard definition yet, which is why a reading map across both fields is useful.

What is the difference between a fitness function and an eval?

In practice, very little. A fitness function is an automated check that an architecture still has a property such as low coupling or acceptable latency. An eval is an automated check that an AI feature still produces acceptable output. Both run in CI and both exist to stop a property from drifting unnoticed.

In what order should I read these three books?

Software architects should read AI Engineering first, since they already know the other two topics. AI engineers should read Fundamentals of Software Architecture, then The Hard Parts. Read The Hard Parts after Fundamentals in either case, because it builds on it.

8. Sources

The mapping between the books is our own reading, not something the authors published. Chapter names and publication dates checked against the publisher's listings on November 16, 2026. No affiliate links.

Found this useful? Share it.

Share:

Related Articles