Forward deployed engineering · Interview prep
Work through real interview questions, learn the engineering behind the answers, and practise on systems you can build and debug.
No card required. Free answers in every topic.
A free lesson to begin · 25 minConnect a software capability to a working customer workflowUpdated · 117 questions added or revised in the last 30 days
17 scenarios to find what needs work. Get a study plan linked to questions and lessons.
Which role are you actually going for?
34 missions in your browser: a support agent that follows instructions hidden in a ticket, a webhook that invoices twice, a model that changed under the same name. Read the traces, watch the metrics move, edit the config, restart the service, and prove the fix from the system’s own state. Nothing to install.
4 free · 3 with a free account · 27 Premium · 20–45 minutes each
From the FDE Lab’s incident response capstone. Inspect the service map, trace the billing failure, and work through recovery in the terminal.
THE FDE VIDEO LIBRARY
Follow a course lesson by lesson, watch a concept take shape, or work through an interview answer. Every video sits beside its written explanation.
Swipe to explore the previews →
9:27Course series · 30 videos
Start with lesson 1
Watch this lesson
9:50Course series · 30 videos
Start with lesson 1
Watch this lesson
3:21Concept explainer
Watch the explanation, then work through the page
Watch this lessonFree samples and Premium lessons. Videos follow the access requirements of their written pages.
Looking for your next FDE role?Explore openings by company, region and work style.
Browse FDE jobsGrounded in real Forward Deployed Engineer interview loops and written to a senior-engineer editorial bar.
ACCURACY YOU CAN AUDITEvery code sample, SQL query and worked number here is executed and checked before it ships, then re-audited. Find an error we can verify and we add a day of full access to your account.How we check, and what counts →FDEInterviews brings interview preparation and ongoing learning into one place. Keep your skills current with courses and labs, follow research and open-source projects, and explore your next FDE role.
FDE FIELD NOTES
Free notes on AI agents and the engineering behind customer deployments. We aim to send a useful concept each week, with clear examples to try in your next build.
Get the free newsletterNo account needed. Unsubscribe anytime.
Join 1,000+ readers.
Including people at Anthropic, OpenAI, Google, Meta, NVIDIA, Microsoft, Amazon and Sarvam.
Resist the urge to touch the prompt. A data refresh changes retrieval, not generation — so measure retrieval before you theorize.
# golden set recall, before touching anything
hits = retriever.search(q, k=20)
print(recall_at_k(hits, gold)) # 0.31 — was 0.86Recall cratered. The refresh re-ran ingestion but skipped re-embedding, so the index points at stale vectors.
408 multiple-choice scenarios across all 11 topics, easy to staff-level. Every answer explains itself and links into the worked material, and the questions rotate so a retake is a fresh test.
A RAG answer is wrong. Which single question splits the failure into its two bug classes?
If the passage never came back, no prompt work will help: fix retrieval. One log lookup separates the two bug classes…
Rate your recall after each answer. Your topic mastery updates as you study, and your next session starts with a suggestion based on the areas that need more practice.
Trace an LLM request from tokenization to the generated response. Each stage names the mechanism, with annotations that connect the explanation to the numbers. 283 answers, concepts and lessons now carry one.
5 courses and 179 lessons take you from your first discovery call to deployments behind an air gap. Work through the customer decisions alongside the engineering.
Each work sample is a small broken thing from a real deployment, handed to you the way a customer hands it over: three symptoms in one ticket, no list of causes. You get the fixtures, a hostile server or a seeded database, and a grader you run locally. Nothing to install, no API key, no account on anything, and every dataset is synthetic and written for the exercise.
The tests going green is the smaller half. Each pack ends with the questions an interviewer asks next, and a reference solution that explains what was being scored rather than only showing the answer. This is the closest thing on the site to the job itself.
THE RESEARCH LIBRARY
291 AI papers explained for engineers. Follow the mechanism, check the results, and connect the research to the systems you build.
Explore the papersFree summaries · Diagrams · Experiments
The FDE reading room
74 pages · 15 diagrams · PDF + tested companion
Evaluate LLMs, retrieval and acting systems. Build datasets, validate judges, measure repeated-run reliability, gate releases and monitor production, with a tested Python/SQLite companion and editable worksheets.
139 pages · 35 diagrams · PDF + tested companion
Build the layer that turns a model into a system a customer can run: the loop, instructions, context and caching, tools and MCP, skills, verification, permissions, credentials, sandboxes, cost, the customer eval and self-improving harnesses, with the 2026 evidence beside each decision and a tested Python companion.
89 pages · 18 diagrams · PDF + tested companion
Turn customer requirements into PRDs, specifications, BDD examples, agent-built code and verified releases. Covers skills, MCP, code property graphs, evals, AgentGym and controlled improvement, with a tested Python/SQLite companion.
105 pages · 29 diagrams · PDF + executable companion
Connect tools, agents, interfaces and commerce through MCP, A2A, UCP, ACP, AP2, x402 and MPP. Learn delegated authority, recover uncertain effects and run a tested Python companion with a real local MCP track.
107 pages · 25 diagrams · PDF + executable companion
Build the data layer behind AI agents: transactional state, vector retrieval, knowledge and context graphs, ontologies, feature stores, catalogs, Open Knowledge Format, MCP and recoverable actions. Includes tested Python and PostgreSQL exercises.
177 pages · 32 diagrams · PDF + reference code
Design agents that act for one person inside an organization: delegated identity, a deterministic gatekeeper, exact-action approvals, memory and recall, voice and messaging surfaces, decision models, and a tested reference build on Google ADK.
66 pages · 20 figures · PDF + calculation companion
Understand vectors, probability, evaluation, gradients, capacity and data visualisation through small worked examples. Includes exercise solutions and runnable calculators.
108 pages · 21 diagrams · PDF + operating toolkit
Establish the mandate, hire and develop the team, govern customer engagements and reconcile the economics. Four worked cases and 24 editable operating forms.
96 pages · 22 diagrams · PDF + Python companion
Design, ship, and operate AI systems in customer environments. Four worked engagements, from permission-aware retrieval to an offline deployment, with nine executable lab tracks.
PAID REMOTE INTERNSHIP
Create useful content and bring students together while learning how AI reaches real customers. Join FDE Interviews as a Community & Campus Intern.
Fully remote · US$25 per week · Outcome-based
Explore the internship ↗
From the blog
Project reviews, worked examples and decisions behind customer deployments.
9 min read
7 min read
10 min read
6 min read
5 min read
5 min read
Not the role in general. The loop: the decomposition round, the practical builds, the applied LLM work, and the customer rounds that decide more offers than anyone admits. Free for premium members on release.
What is inside the book →Every answer, concept and course, every paper breakdown, all published video lessons, hands-on FDE Lab missions, Premium PDF guides and companion files, the full practice-test bank and work-sample downloads. Referral Premium excludes guide PDFs and their companion files.
6 months · One payment · No auto-renewal
Study alongside free video lessons.
launch price · regularly ₹4,000 / $50
The question bank
11 topics · Curated journeys
Each path moves from the fundamentals to harder interview problems.