Applied LLM Engineering — From Fundamentals to Production

AI-Generated Content

AI-generated content may contain errors. Always verify against official sources.

Applied LLM Engineering — From Fundamentals to Production

A practitioner's guide to building reliable LLM pipelines — from understanding how they work to shipping production systems.

---

What You Will Learn

By the end of this course you will be able to:

• Explain how LLMs work at the token and sampling level
• Call any LLM API (OpenAI, Anthropic, Google Gemini, open-source) confidently
• Write production-quality prompts and evaluate them systematically
• Build Retrieval-Augmented Generation (RAG) pipelines from scratch
• Chain LLM calls using LangChain
• Create autonomous agents with tool-calling
• Fine-tune and deploy open-source models
• Evaluate and monitor LLM systems in production

---

Course Map

#	Chapter	Core Skill
1	LLM Foundations	How models work, tokenization, model landscape
2	Prompt Engineering	Zero/few-shot, chain-of-thought, structured output
3	LLM APIs	OpenAI, Anthropic, Gemini, open-source via Ollama
4	RAG	Embeddings, vector stores, retrieval pipelines
5	Chains	LangChain LCEL, multi-step workflows
6	Agents	Tool-calling, ReAct, agent frameworks
7	Fine-Tuning	LoRA, QLoRA, instruction tuning
8	Evaluation	RAGAS, LLM-as-judge, tracing
9	Deployment	FastAPI, Docker, cost optimisation

---

Official Documentation References

Every chapter links to primary official sources:

• 📘 OpenAI Platform Docs
• 📘 Anthropic Claude Docs
• 📘 Google Gemini API Docs
• 📘 Hugging Face Docs
• 📘 LangChain Docs
• 📘 Ollama Docs

---

Prerequisites

• Python 3.10+ — comfortable with functions, classes, and pip
• Basic understanding of REST APIs
• No ML background required

Start Here

Head to 1.1 What is an LLM? to begin.