Intermediate · Free course
Generative AI Engineering
Build with real LLM APIs: prompting, RAG, tools and evaluation.
90 lessons3 videos so far40 min of videoNotes for every lesson
What you’ll learn
- Tokens, models and sampling
- Prompting and structured output
- Embeddings and RAG
- Tool use, agents and evals
0 of 90 complete
Curriculum
- 1What an AI engineer buildssoon
- 2Python setup for AI engineeringsoon
- 3HTTP and JSON for LLM APIssoon
- 4Tokens and tokenizers in codesoon
- 5Your first LLM API callsoon
- 6Sampling and decoding parameters14 min
- 7Multi-turn chat and the context windowsoon
- 8Streaming responsessoon
- 9Errors, retries and rate limits13 min
- 10Async, concurrency and batch calls13 min
- 11Providers, model families and abstraction layerssoon
- 12Anatomy of a good promptsoon
- 13Zero-shot and few-shot promptingsoon
- 14Reasoning prompts and reasoning modelssoon
- 15Prompt chaining and decompositionsoon
- 16Prompt templates and versioningsoon
- 17Self-critique, refinement and advanced patternssoon
- 18Hallucinations and how to reduce themsoon
- 19Structured output 1: JSON mode and parsingsoon
- 20Structured output 2: schemas, Pydantic, constrained decodingsoon
- 21Extraction and classification with LLMssoon
- 22Tool calling 1: your first function callsoon
- 23Tool calling 2: multiple and parallel toolssoon
- 24Building useful toolssoon
- 25Embeddings in codesoon
- 26Similarity measures and semantic searchsoon
- 27Vector search from scratchsoon
- 28ANN indexes: HNSW, IVF and PQsoon
- 29Vector databasessoon
- 30Choosing an embedding modelsoon
- 31RAG 1: the basic pipelinesoon
- 32RAG 2: loading and cleaning documentssoon
- 33RAG 3: chunking strategiessoon
- 34RAG 4: grounded answers and citationssoon
- 35RAG 5: keyword search and BM25soon
- 36RAG 6: hybrid searchsoon
- 37RAG 7: rerankingsoon
- 38RAG 8: query rewriting and expansionsoon
- 39RAG 9: advanced retrieval patternssoon
- 40RAG 10: GraphRAG and structured datasoon
- 41RAG 11: evaluating retrievalsoon
- 42RAG 12: evaluating generationsoon
- 43RAG 13: production RAGsoon
- 44Agents 1: the agent loop in codesoon
- 45Agents 2: planningsoon
- 46Agents 3: memory and context engineeringsoon
- 47Agents 4: agentic RAGsoon
- 48Agents 5: workflows vs agentssoon
- 49Agents 6: multi-agent systemssoon
- 50MCP 1: concepts and clientssoon
- 51MCP 2: build an MCP server in Pythonsoon
- 52Agents 7: coding and computer-use agentssoon
- 53Agents 8: human-in-the-loop and reliabilitysoon
- 54Agent frameworks comparedsoon
- 55Evals 1: building test setssoon
- 56Evals 2: metrics and code graderssoon
- 57Evals 3: LLM-as-judge done rightsoon
- 58Evals 4: agents, benchmarks and online testssoon
- 59Observability for AI appssoon
- 60Prompt regression tests and optimisationsoon
- 61Guardrailssoon
- 62Prompt injectionsoon
- 63Defending against injection and data leakagesoon
- 64Responsible AI and compliance in practicesoon
- 65Cost optimisationsoon
- 66Latency optimisationsoon
- 67Cachingsoon
- 68Open-weight modelssoon
- 69Local inferencesoon
- 70Quantisationsoon
- 71When to fine-tunesoon
- 72Preparing fine-tuning datasoon
- 73Fine-tuning hands-on with Hugging Facesoon
- 74LoRA and QLoRAsoon
- 75Preference tuningsoon
- 76Serving models at scalesoon
- 77Vision: images in promptssoon
- 78Audio and voice appssoon
- 79Image generation and multimodal RAGsoon
- 80Building a chat UIsoon
- 81Deploying an AI appsoon
- 82Production concernssoon
- 83LLMOps lifecyclesoon
- 84Interview prep 1: core GenAI conceptssoon
- 85Interview prep 2: GenAI system designsoon
- 86Interview prep 3: coding rounds and portfoliosoon
- 87Project 1: docs Q&A botsoon
- 88Project 2: coding helper agentsoon
- 89Project 3: research assistantsoon
- 90Capstone: ship a production AI appsoon