Fundamental knowledge, advanced
Martin Kleppmann & Chris Riccomini, 2026-02. Paid / library access
Build a vocabulary for correctness and performance tradeoffs across data systems.
Used in chapter 12: Defend the work, chapter 20: Protect the invariant before optimizing recovery time
How to use it
- Focus your reading
- Second edition: chapters 5, 6, 8, 9 and 12 on evolution, replication, transactions, failures, and streams.
- Put it to work
- Defend one consistency contract: stale reads allowed, invariants protected, and recovery under a partition.
- Before you start
- Use selected chapters as a reference. This is not a requirement to build a distributed database.
Publisher description and contents checked; full paid text not reviewed. Reviewed 2026-10-31
Fundamental knowledge, practitioner
Gregor Hohpe & Bobby Woolf, 2003-10. Paid / free selected excerpts
Design message flows across systems with different schemas and failure semantics.
How to use it
- Focus your reading
- Prioritize message construction, routing, transformation, and endpoints; look for correlation and aggregation.
- Put it to work
- Draw a multi-system workflow with correlation IDs, timeout behavior, and handling for late or duplicate messages.
- Before you start
- Examples use older technology; transfer the patterns and evaluate their fit to your stack.
Publisher description and contents checked; selected author excerpts checked. Reviewed 2026-10-31
Discovery & scoping, foundation
Rob Fitzpatrick, Author book page. Paid
Reduce leading questions and false enthusiasm during discovery.
Used in chapter 3: Decompose and scope
How to use it
- Focus your reading
- Read for biased feedback, concrete customer conversations, and evidence of commitment.
- Put it to work
- Rewrite five hypothetical questions and identify the next observable commitment.
- Before you start
- Choose this when interviews are new to you; experienced interviewers can use the free YC talk as a refresher.
Dedicated book page checked; full paid text not reviewed. Reviewed 2026-10-31
Discovery & scoping, practitioner
Peter Block, 2023-05. Paid / library access
Make expertise usable when the sponsor, operator, and engineer want different things.
Used in chapter 2: Discover the real problem, chapter 18: A pilot can be technically good and operationally abandoned
How to use it
- Focus your reading
- Read for contracting, mutual expectations, resistance, and ownership of implementation.
- Put it to work
- Create an engagement agreement: sponsor access, user access, decisions, responsibilities, and exit conditions.
- Before you start
- Book recommendation based on publisher scope; this guide does not reproduce its methods or claim a full-text review.
Indexed publisher description checked; direct access blocked; full paid text not reviewed. Reviewed 2026-10-31
Mindsets & judgment, advanced
Will Larson, 2021. Paid / free companion guide
Move from solving visible tasks to choosing consequential work across team boundaries.
Used in chapter 1: Understand the role
How to use it
- Focus your reading
- Start with the free “Work on what matters” guide; use the book for archetypes and organizational alignment.
- Put it to work
- Write a decision memo: why this problem, why now, who must agree, and what you will stop doing.
- Before you start
- The free guide is a starting point; buying the book is optional. FDE and Staff Engineer are different role dimensions.
Author book page and linked free guide checked; full paid text not reviewed. Reviewed 2026-10-31
Agents & tool design, practitioner
Chip Huyen, 2024-12 · copyright 2025. Paid / library access
Use one broad reference to connect evaluation, RAG, adaptation, and inference economics.
Used in chapter 17: Lower token cost can raise cost per successful task
How to use it
- Focus your reading
- Chapters 3–4 for evaluation; 6 for RAG/agents; 7 for fine-tuning; 8–10 for data, serving, and feedback.
- Put it to work
- Build a decision table: prompt change, retrieval change, fine-tuning, or no model; include the evidence required.
- Before you start
- Read the chapter matching your bottleneck. Current model capabilities and APIs need separate checking.
Publisher description and contents checked; full paid text not reviewed. Reviewed 2026-10-31
Production engineering, practitioner
Google SRE contributors, 2018. Free
Make a reliability target drive an actual decision.
How to use it
- Focus your reading
- Read good-events/total-events indicators, stakeholder agreement, and error-budget policy.
- Put it to work
- Write one user-centered indicator and the action taken when its budget is exhausted.
- Before you start
- Prefer this implementation chapter if you already know SLI/SLO definitions; the original chapter remains in citations.
Content checked. Reviewed 2026-10-31
Production engineering, advanced
Alec Warner & Stepan Davidovic, with contributors · Google SRE, 2018. Free
Distinguish a small deployment from a useful release experiment.
Used in chapter 7: Engineer for production, chapter 20: Protect the invariant before optimizing recovery time
How to use it
- Focus your reading
- Read representative canaries, comparison signals, and automated rollout decisions.
- Put it to work
- Define cohort, exposure, observation window, rollback trigger, and how low traffic affects detection.
- Before you start
- A canary that sees no representative failures cannot establish that the full rollout is safe.
Content checked. Reviewed 2026-10-31
Productization & leadership, advanced
Matthew Skelton & Manuel Pais, 2025. Paid / library access
Design a handoff that respects team capacity and cognitive load.
Used in chapter 11: Generalize and hand off, chapter 19: Productize the repeated constraint
How to use it
- Focus your reading
- Read for team boundaries and interaction modes: collaboration, facilitation, and service consumption.
- Put it to work
- Name when an embedded engagement ends and how the customer team gets a supported service afterward.
- Before you start
- An organization-design lens; avoid copying a team chart without understanding the work.
Author book page checked; full paid text not reviewed. Reviewed 2026-10-31
Productization & leadership, advanced
David Challoner et al. · Google SRE, 2018. Free
Count the work a successful deployment creates for its maintainers.
Used in chapter 11: Generalize and hand off, chapter 17: Lower token cost can raise cost per successful task, chapter 19: Productize the repeated constraint
How to use it
- Focus your reading
- Read identifying, quantifying, and prioritizing toil; inspect the case examples.
- Put it to work
- Record support time per successful task and compare automation cost with recurring manual effort.
- Before you start
- Google’s operational-work target is contextual, not a required FDE staffing ratio.
Content checked. Reviewed 2026-10-31