The hands-on course for Forward Deployed Engineers
Discovery calls, messy data, flaky APIs, Claude-powered agents, evals and production incidents. Practice every part of the job through six realistic customer engagements, writing real Python in your browser.
No installs · No API key needed · Lifetime access
def guarded_execute(block, approver, audit): if needs_approval(block.name, block.input): request = {"summary": describe_action(...)} try: approved = approver(request) is True except Exception: approved = False # fail closed
Forward Deployed Engineers sit between the customer and the code. They find the real problem, build the solution in the customer's environment, and own it until it works. This course teaches that job end to end.
moving into Forward Deployed, solutions or customer-facing engineering roles.
who want to build and ship the systems, not just design them.
who can build a demo and now need to get it working inside an enterprise, safely.
Each module puts you inside a realistic customer with real constraints, messy data and a stakeholder waiting for results.
Profile, clean and reconcile messy CRM and billing exports
Integrate a paginated, rate-limited partner API and webhooks
Answer the COO's questions with SQL, cohorts and a metrics layer
Build an injection-resistant ticket classifier with Claude
Ship a support agent with RAG, approvals, evals and production monitoring
Capstone: scope, build, evaluate and present an exception-triage agent
Lesson on the left, code on the right, instant feedback.
Detailed written lessons built from real deployment patterns: discovery calls, messy data, flaky APIs, LLM evals, production incidents.
A full Python editor runs in your browser, including a simulated Anthropic SDK. No installs, no API key, no setup.
Every exercise has automatic checks and step-by-step hints, plus an AI tutor powered by Claude when you're stuck.
From your first customer call to running Claude-powered systems in production. Module 1 is free.
9 lessons · 2.3 hours · Free
What forward deployed engineers actually do, how the role differs from SWE, solutions engineering and consulting, and the operating loop you'll run on every engagement.
12 lessons · 3.8 hours
Turn vague asks into scoped, measurable engagements: discovery calls, stakeholder maps, honest baselines, requirements extracted with Claude, and change requests handled without saying no.
13 lessons · 4.8 hours
Customer data is never clean. Profile it, normalize it, reconcile systems of record, deduplicate entities, label free text with Claude cost-effectively, and report quality so problems get fixed at the source.
11 lessons · 4 hours
Connect to customer systems that paginate, rate-limit, time out and change without warning. Build clients, webhook receivers and sync jobs that survive production, and use Claude to speed up integration work safely.
11 lessons · 4.1 hours
Query customer databases safely and fluently: answer executive questions, use window functions and cohorts, build a metrics layer, guard production data, and put Claude-powered text-to-SQL behind real guardrails.
11 lessons · 4 hours
The core of modern FDE work: the Anthropic SDK in depth. Multi-turn state, production prompts with injection defenses, streaming, errors and refusals, prompt caching and real cost, and routing requests across models.
9 lessons · 3.7 hours
Give Claude access to customer systems. Define strict tools, write a traced agent loop, answer from documents with checked citations, gate risky actions behind human approval, and know when not to build an agent.
11 lessons · 4.3 hours
Prove your LLM system works before the customer finds out it doesn't. Build eval harnesses, calibrate LLM judges, evaluate agent trajectories and retrieval, gate releases on regressions and noise, and model cost against quality.
11 lessons · 4.1 hours
Deploy into environments you don't control. Map the customer's environment, validate config and protect secrets, log every Claude call safely, alert on what matters, degrade gracefully, debug an incident from logs alone, and hand off with runbooks and a blameless post-mortem.
9 lessons · 4.4 hours
Run a full simulated engagement for NorthStar Logistics: scope from discovery notes and a measured baseline, clean and join their messy data, design and build a guarded exception-triage agent, evaluate it and make the ship call, then turn a pilot into fact-checked numbers for the executive readout.
Who I am
I'm a seasoned AI architect. I worked at Amazon, and today I'm an FDE Architect at an AI startup, working in the Forward Deployed Engineering model: embedded with customers, from the first discovery call to the system running in their production environment.
I've shipped AI-based solutions to enterprise customers, and I've learned that the hard parts are rarely the model. They're the messy data, the integration nobody documented, the security review, the evaluation that proves it works, and the readout that earns the next phase.
I built FDE Playbook to teach exactly those parts, the way I wish I'd learned them: through realistic customer engagements and code you write yourself.
Start free. Upgrade when you're ready.
Module 1 is free. No credit card needed to start.
No. Lessons, the code editor and the Python runtime all run in your browser. The code you write uses the same Anthropic SDK calls you'd use in production.
Comfort with functions, lists and dictionaries is enough to start. Module 3 strengthens the data-wrangling skills the rest of the course relies on.
Not for the course. Exercises run against a built-in simulator of the SDK, so you can learn without paying for API calls. Your code works against the real API unchanged.
About 40 hours of lessons and exercises. Most people finish in 4 to 8 weeks at a few hours a week. Access never expires.
Yes. Module 1 is free, with no credit card needed.
Module 1 is free and takes about two hours. No setup, no credit card.
Start Module 1 free →