The newspaper of the Forward Deployed Engineer
Books and courses that help you move into FDE work, in the order to take them.
Marty Cagan writes for product managers, but his thinking names the trap FDEs fall into most often: shipping every feature on the customer's list without knowing what problem has been solved.
09/10/2026
The customer system you are about to wire an LLM into almost certainly has no tests. Michael Feathers explained how to handle that more than twenty years ago.
A client API that suddenly hangs for 30 seconds per call can drain your service's threads in just 5 seconds. This 376-page book shows how to break that chain of failures at design time.
The course is free and available in many languages. Its first three chapters can be read as the order of work at a client site: set a baseline first, inspect the data next, fine-tune last.
At consulting firms, a draft can come back with a single instruction: "make it Minto". Forward deployed engineers, who write for client leadership, would do well to pick up the same habit.
When a client asks why a model gave the wrong answer, an FDE who has written backprop and a tokenizer by hand can explain it. Someone who only knows how to call the API cannot.
If you are about to read a client's database, redesign its tables and write safe queries, you can practise all three in a free course taught by David J. Malan and Carter Zenke.
Melissa Perri wrote it for product managers, but the readers who need her 2018 book most are engineers who sit in a client's office and decide every week what to build next.
Ten years on, the remake has moved to Python. For developers who want to become FDEs, the course's real value is that it teaches you a model well enough to explain it to a client.
On a client site, the hardest part is usually not the pipeline but the silent room after the demo. Peter Block wrote this book to teach you how to handle that room.
Jeremy Howard's free course has you running a model on real data in the first lesson. The theory comes later.
American clients tend to praise first and criticise second, German clients say it straight, and Japanese clients convey a lot through what they leave unsaid. Erin Meyer's book helps FDEs hear all three correctly.
Put the syllabus next to Anthropic's own FDE job posting and it becomes clear which modules deserve close study, which you can skim, and which parts of the work you will have to learn outside the course.
First published in 2018, the book by Nicole Forsgren, Jez Humble and Gene Kim gives you hard numbers for the question clients ask most often: how fast and how reliably does our team get changes into production?
At a client site there is no team lead to speak up for you. The book by David Thomas and Andrew Hunt teaches the habits you need when you are on your own.
More and more data systems keep their data in object stores, not on a server's local disk. The old trade-offs are all still there, just in different places, and this book helps you find them.
Last quarter's figures change even though nobody touched the code. The cause is usually a modelling decision that Ralph Kimball and the Kimball Group named and explained long ago.
Data tools have changed several times since 2022. The six "undercurrents" that Joe Reis calls the most useful part of the book still give a forward deployed engineer the right questions for a first customer meeting.
The most useful part of this system design interview book is not its 30 sample designs. It is the four-step answer framework, and FDE candidates should treat it as a guide to their first working session with a client.
The next time a client asks for "just one small extra feature", you do not have to choose between saying yes and pushing back. You can ask the right question instead.
The syllabus puts the first question you ask a customer ahead of the first design you draw. Candidates who prepare mainly with LeetCode tend to skip that step.
The DataTalks.Club course keeps going after the chatbot answers its first question. Two of its modules cover how to measure and watch the system once it is running.
Three Google engineers wrote a book that says very little about writing code. It is about what happens to code after you leave the project, and a forward deployed engineer has to answer that question at every customer.
The most valuable part of the course is not the six modules. It is the last three weeks, when you build a complete pipeline yourself.
The course launched in 2023, so its techniques are no longer new. What it teaches best still holds: how to think through an entire LLM project, from choosing a use case to deployment, which is exactly what an FDE has to do in front of a client.
Most people talk about reflection, tool use, planning and multi-agent systems. The module on error analysis is closer to what a Forward Deployed Engineer actually does at a customer site every day.
Chip Huyen's book is long, but the problems an FDE meets every day at a customer site fit into four chapters: data, training data, model updates and infrastructure.
Chip Huyen says plainly that this is not a tutorial book. That admission is the reason a forward deployed engineer should read it before any step-by-step coding guide.
Once a client stops telling you the truth, no log or evidence will save the meeting. This book shows what to do first.
Four units, three frameworks, 3–4 hours a week. The course's real value is that it makes you put your agent on a benchmark and accept a pass mark far lower than the quiz's.
Three consultants broke trust into four variables. The one in the denominator decides whether a customer listens to you.
Maven's highest-grossing course has taught more than 5,000 engineers and PMs, and its central argument is that error analysis matters more than any eval tool.
A discovery session full of compliments is a failed discovery session. This slim book shows how to spot that before you write your first line of code.
