The LLM Engineering Roadmap

What a large language model really does, why calling it from a demo is easy and shipping it is hard, and how this course runs code in three languages.

Tokens, embeddings, attention, the context window, sampling, training, reasoning and hallucination, each built by hand with small numbers.

A model call is an HTTPS request with JSON. Build it by hand, read the reply, keep keys safe, classify errors, set timeouts and hide providers behind one interface.

Treat prompts like code: write them as specifications, keep user data apart from instructions, build them with functions, version them and test them.

Make a model return data a program can trust: JSON mode, schemas, validation, repair loops, honest unknowns, and one typed-call function in three languages.

Show the first words in under a second: server-sent events, assembling a streamed reply, cancelling, partial JSON and streams that break midway.

What every model call costs and how long it takes, then caches, rate limits, retries with jitter, fallbacks, circuit breakers and concurrency, built by hand first.

How a model asks your code to act: the tool-call loop built by hand, schemas, validation, errors, parallel calls, dangerous tools, MCP and a safe tool runner.

Give a model the right pages of your own data: keyword search, vector search, chunking, a working pipeline, numbered citations and honest refusals.

How to know your LLM feature works, and keep knowing it: golden sets, scorers, judges, regression gates in CI, fakes, recordings and production feedback.

Defend an LLM app: threat models, prompt injection, untrusted output, data leaks, PII, guard models, least-privilege tools and a red-team suite.

See what your LLM app is doing: safe logs, traces, metrics, alerts, user feedback and incident drills, built by hand and then with real tools.

Put the pieces together: the layers of a real LLM backend, its API, stored conversations, queues, tenants, routing, config, deployment and a service skeleton.

Pick a model with evidence, not hype: scorecards, open vs hosted, quantisation, local servers, fine-tuning, distillation and honest reading of benchmarks.

Answer the questions LLM interviewers ask, design two LLM systems on a whiteboard, solve the coding-round problems, then build a production-grade assistant.