LLM Engineering
Build with large language models: tokenization, embeddings, RAG, fine-tuning, evals, agents and production LLM systems.
What Is LLM Engineering?
The discipline of building systems on top of large language models — and where it differs from classic ML.
LLM Architecture Review
Decoder-only transformers, context windows, and what happens inside a single generation call.
Working with LLM APIs
Chat completions, parameters, streaming and structured outputs — the daily API toolkit.
Prompting Systems at Scale
Turn prompts into versioned, testable system components — not strings in code.
Tokenization & Context Management
Budget tokens, count costs, and fit big context into small windows.
Embeddings & Semantic Search
Turn text into vectors and search by meaning — the retrieval backbone of LLM apps.
RAG Engineering
Design retrieval-augmented generation that is reliable: chunking, indexing and prompting.
Advanced RAG
Hybrid search, reranking, query rewriting and contextual compression for better answers.
Fine-Tuning LLMs in Practice
LoRA, QLoRA and instruction tuning — adapt open models when prompting is not enough.
Function Calling & Structured Outputs
Let the model call your tools with typed arguments — reliably.
Building LLM Agents
Connect models to tools and loops: the agent pattern in production.
Context Engineering & Memory
Manage what the model sees: conversation memory, summarization and caching.
Evaluating LLM Systems
Build evals that catch regressions: golden sets, LLM-as-judge and human review.
Guardrails & Safety for LLM Apps
Prompt injection, moderation, PII redaction and output filtering.
LLM Serving & Inference
Serve open models: vLLM, quantization, batching and GPU memory management.
Cost Optimization for LLM Apps
Cut spend without cutting quality: caching, smaller models, and smart routing.
Prompt Versioning & Observability
Trace every request: prompt, tokens, latency, cost and outcome.
Building a Production Copilot
Assemble everything into a real product: an assistant grounded in your data.
LLM Production Case Studies
Learn from real systems: search copilots, support agents and coding assistants.
The LLMOps Tooling Landscape
Survey the ecosystem: frameworks, eval platforms, observability and serving.
LLM Engineering Roadmap
Synthesize the course into a plan: build, ship and scale LLM products.