Architecture
What is RAG (Retrieval-Augmented Generation)?
Retrieval-augmented generation, or RAG, lets an AI answer from your own records instead of only what it learned in training. At answer time the system retrieves the relevant facts from your documents or database and hands them to the model, so the reply is grounded in your prices, your policies, and your products.
01In practice
02Keep reading
The terms around it.
- Architecture
MCP (Model Context Protocol)
The standard that lets an agent use your tools through one interface.
Read the definition - Architecture
Fine-tuning
Training a model further on your own examples.
Read the definition - Architecture
Vector Database
How an AI stores and searches meaning, not keywords.
Read the definition - The install
Mapping Call
The first call. Bring one process you still run by hand.
Read the definition - The install
Operating Map
How the work actually moves, written down before scope is set.
Read the definition - The install
Set Price in Writing
One price, after mapping, that moves only if the scope does.
Read the definition
Your first department
Bring the process you still run by hand.
One mapping call. You leave with the system design, the price in writing, and the date it goes live. The first build lands in two weeks.
Glossary
Every term, on one page.
The words on the first call, the controls, and the pieces underneath them.
Back to the glossary