AI Technology Knowledge Base

RAG, LangChain, MCP, and more — explained simply with real examples

RAG (Retrieval-Augmented Generation) is a technique for making an AI model answer using your own documents.

It matters because it lets an AI stay accurate about information it was never trained on.

See a real local RAG implementation →

Five AI technologies, explained

Accurate, approachable explainers of real, widely-used technologies — not marketing claims.

The foundation

LLMs (Large Language Models)

A neural network trained to predict the next word across huge amounts of text — which, done well at scale, is what lets it write, reason, and hold a conversation. Comes with real limits: a training knowledge cutoff, a finite context window, and no built-in fact-checking.

Read the full guide →

Applications — what you build with an LLM

AI Agents

A model that plans a sequence of steps, uses tools to take real actions (edit files, run commands, call APIs), observes the results, and decides what to do next — repeating that loop until a task is done, not just answering one question at a time.

Read the full guide →

Local Knowledge Bases (RAG + LangChain)

How to give a model accurate answers about your own documents without retraining it: retrieve relevant chunks (RAG), then hand them to the model alongside your question. LangChain is the most common framework for building this kind of pipeline.

Read the full guide →

How Claude Code itself is extended

Skills

A packaged, specialized instruction set that an AI agent loads only when actually relevant to the task at hand, instead of every conversation carrying the full weight of every capability it might ever need.

Read the full guide →

MCP (Model Context Protocol)

An open standard for connecting AI models to external tools and data — file systems, databases, browsers — through a common protocol instead of a custom integration per model/tool pair. Think USB for AI-to-tool connections.

Read the full guide →