Skip to content

Explore ChinFDE

FDE course

Pair with AI throughout: AI drafts the code and SQL, and you learn to read, judge, verify and sign off on delivery. Each chapter ends in a verifiable lab output.

A systematic course for Forward Deployed Engineers: customer data integration, cloud and delivery engineering, on-premises deployment, RAG and agents, evaluation-driven delivery and on-site methods. 13 chapters, 128 lessons, 9 labs and 60 self-check questions, with an AI tutor.

Getting started: How to take this course and what to prepare

  • Intro How to Use This Course

    6 lessons · about 1 h

    See the whole map before the details. This course teaches you to pair with AI on FDE work. AI drafts the code, SQL and configuration. You own the judgment and the sign-off.

Role and positioning: What an FDE is and isn't, and the path to becoming one

  • Chapter 0 Meet the FDE: A Widely Misunderstood Role

    11 lessons · about 1 h

    Before you write your first line of code, you need to know where you are going. This chapter breaks down what the FDE role actually does day to day, where it sits relative to developers, ops and pre-sales, where this course's knowledge structure comes from, and how to spend 16–20 weeks.

Foundations: Computing and networking basics, programming and integration, customer data integration

  • Chapter 1 Computer and Network Fundamentals: Opening the Black Box

    12 lessons · about 2 h

    The most common weakness among people who have spent a few years in business software development is not writing code. It is being unable to explain what is actually happening inside a server, or beneath a single request. On a customer's site, that gap shows up on day one. This chapter is a preparatory track for learners starting from zero. It covers one thing only: what goes on inside one server, one network request, and one code repository.

  • Chapter 2 Programming and Data: The Craft of the FDE

    10 lessons · about 7 h

    The code an FDE writes most often is not algorithms but code that "reliably connects two systems." The focus of this chapter is not syntax but five engineering mechanisms that decide whether a project lives or dies: authentication, pagination, timeouts, retries, and idempotency.

  • Chapter 11 Data Integration: Turning Customers' Dirty Data into a Trustworthy Foundation

    11 lessons · about 7 h

    Numbering note: this chapter was added later, so its number comes last, but the learning order follows Chapter 2 directly. Chapter 2 teaches you how to connect two systems. This chapter handles what happens after they are connected: where the customer's data comes from, whether definitions are consistent, whether the same customer is the same entity across three systems, and how yesterday's and today's data reconcile. In FDE projects, most of the workload and most of the delays happen at this step.

Cloud and delivery engineering: Cross-cloud models, IAM and networking, containers and IaC, observability and cost

  • Chapter 3 The Cloud Mental Model: Four Clouds, One Language

    9 lessons · about 1 h

    This chapter is where the course's "cloud-neutral" promise is delivered. We align the core products of AWS, Google Cloud, Alibaba Cloud, and Tencent Cloud by capability into one comparison table, and then teach you how to choose among five compute forms. The basis for the choice is engineering constraints, not product names.

  • Chapter 4 Permissions and Networking: Hard Skills You Need on Day One

    11 lessons · about 5 h

    Customers will not hand you an admin account. In the first few days on site, over half of the errors you hit trace back to "permission not granted" or "a network hop was blocked." This chapter gives you two troubleshooting procedures that you can run directly and that work across every cloud.

  • Chapter 5 Containers and Infrastructure as Code: Making Deliverables Portable

    9 lessons · about 7 h

    Docker solves "it works on my machine." OpenTofu solves "an environment built by hand cannot be reproduced." Together they are how the "cloud neutrality" described in Section 0.6 becomes practical.

  • Chapter 6 Observability and cost: the real work starts after launch

    8 lessons · about 5 h

    Deploying code to production does not mean delivery is complete. One of the core abilities of an FDE is "independently locating an anomaly using only production observability data." The other is "making sure the customer's bill does not spiral out of control in the middle of the night." In the customer's eyes, these two things matter far more than whether your architecture diagram looks polished.

Applied AI and field work: RAG, agents, eval-driven delivery, private and domestic-stack deployment, and how to work on site

  • Chapter 7 Applied AI Engineering: RAG, Agents, and Eval-Driven Delivery

    10 lessons · about 21 h

    Enterprises don't buy "the ability to call a model API." They buy "a system that embeds into business processes and produces reliable results." This chapter covers only engineering, not prompt mysticism. Every principle serves one goal: making the system trustworthy on site.

  • Chapter 12 Deploying in China: private, air-gapped, Xinchuang, and MLPS 2.0

    10 lessons · about 9 h

    Numbering note: this chapter was added later. Read it right after Chapter 7. In AI projects in China, questions like "Can it be deployed in our own data center?", "Can it run with no internet?", "Do we use a domestic model?", and "Will it pass MLPS?" often come up in the very first meeting, and they directly determine the design and the quote. Xinchuang is China's domestic IT stack initiative, and MLPS 2.0 is China's Multi-Level Protection Scheme for information security. This chapter treats these constraints as default conditions, and ends with a lab that deploys the same system on public cloud and on an air-gapped intranet, then compares the results with data.

  • Chapter 8 Field Playbook: From "Can Write Code" to "Can Get It Done"

    11 lessons · about 1 h

    Technical depth determines how fast you can move. Your ability to break down problems and drive them to delivery determines whether the project is renewed or abandoned. This chapter is an operating manual you can print and bring into the customer's meeting room.

Assessment and growth: Certification and job search, the assessment standard, after the course

  • Chapter 9 Free Credits, Certifications, and Job Hunting: Turning Learning into Opportunity

    7 lessons · about 1 h

    This chapter is practical. It covers how to practice at zero cost without surprise charges, which certifications are worth taking and which are a waste of time, and how to prepare for each of the four interview rounds.

  • Chapter 10 Assessment, Graduation and Continued Growth

    3 lessons · about 1 h

    Use this to check whether you are truly at hire-ready level, and to plan the 90 days after graduation. Instructor scheduling and class management are in Appendix I.

See all features