Start here in Generative AI
The most useful guides in this topic, not just the most recent.
Generative AI Guides & Tutorials
Practical guides to LLMs, RAG, agents, prompt engineering and the modern Gen AI stack used in production today.
Generative AI articles
Retrieval Evaluation: Measuring RAG Quality
You cannot improve a RAG system you have not measured. This guide computes hit rate, MRR and nDCG on 40 real questions, then shows two things most tutorials skip: how wide the uncertainty is on a small test set, and why a retriever that looks better may not be better.
Semantic Search vs Keyword Search: What Changed with AI
Keyword search matches the words you typed. Semantic search matches what you meant. Each fails where the other succeeds, and this guide shows it with a test: 40 questions over 148 sections of Python documentation, answered by BM25, by embeddings and by a hybrid of the two.
Vector Databases Explained: Pinecone, FAISS and ChromaDB
A vector database stores embeddings and finds the nearest ones to a query without comparing against all of them. This guide explains the indexes that make that possible, measures recall against work on 35,000 real embeddings, and separates a library, an embedded database, a hosted service and a Postgres extension.
Function Calling and Tool Use in LLMs Explained
Function calling lets a language model ask your code to look something up or take an action. This guide explains the round trip and how to define tools, then puts ten customer requests to a small local model with three tools, and shows the four ways it went wrong and the safeguards that caught them.
Guardrails for LLMs: Keeping AI Apps Safe and On-Topic
Guardrails are the checks around a language model that keep an application safe and on task. This guide explains input, output and action guardrails, then hides four prompt injections in product reviews and measures how many get past a plain prompt, a delimited prompt and a layered defence.
How to Reduce LLM Hallucinations in Production
Language models state wrong things fluently, and in production that becomes a support or legal problem. This guide explains why models hallucinate, then asks a small model twelve questions about an invented shop's policies under four set-ups, from no document to a document with rules and a quote check in code.