Learning path
RAG: from idea to a trusted assistant
Understand how it works, split documents well, build a pipeline and prove it answers correctly.
- Length
- 5 articles · about 37 min
- Level
- Intermediate
- Before you start
- Basic Python for the tutorial. No machine learning background needed.
Best for: Teams building an AI assistant over their own documents
By the end
What you will be able to do
- Explain how retrieval-augmented generation works end to end
- Understand embeddings and choose a chunking strategy for your documents
- Build a small retrieval pipeline without a framework
- Measure retrieval and answer quality with a real test set
- Spot and fix the most common failure modes
The path
Read in this order
Concept · Beginner · 7 min read
RAG explained: how AI answers from your own documents
Retrieval-augmented generation in plain language: how it works, why it reduces made-up answers, and where it can still fail.
You will learn
- RAG retrieves relevant passages first, then asks the language model to answer using only those passages.
- Good retrieval matters more than a bigger model, so most of the work is in how documents are prepared and searched.
Start here →
Concept · Beginner · 6 min read
Embeddings in plain language: how AI measures meaning
The idea behind modern AI search: turning text into numbers so that similar meanings sit close together.
You will learn
- An embedding turns text into a list of numbers that captures its meaning.
- Similar meanings produce nearby numbers, so search works on ideas, not just matching words.
Read →
Concept · Intermediate · 6 min read
Chunking for RAG: how to split documents so AI can find answers
Why chunk size changes everything in a document assistant, and a practical way to choose a strategy.
You will learn
- Chunks that are too large blur many ideas together. Chunks that are too small lose the context needed to answer.
- Start from the natural structure of your documents: headings, sections and paragraphs.
Read →
Tutorial · Intermediate · 12 min read
Build a minimal RAG pipeline in Python
Chunk text, embed it, retrieve the best passages, build a grounded prompt and test it, with no framework.
You will learn
- A working RAG system is a few small functions: chunk, embed, retrieve and build a prompt.
- Chunking with a little overlap keeps ideas from being cut in half.
Read →
Guide · Intermediate · 6 min read
How to evaluate a RAG assistant before you trust it
A five-step test plan to measure whether an AI assistant answers correctly from your documents, and knows when not to answer.
You will learn
- Build a test set of real questions with known answers before launch.
- Measure retrieval and answer quality separately, so you know where problems start.
Read →
When you finish
Where to go next
Next path
Adopt AI in your organization
Learn the language and how AI works first, then check your readiness, run one measured pilot and understand the two big building blocks.
6 articles · 46 min
Next path
Back to basics: the maths and science behind GeoAI, 3D and VR
AI can draft the code, the model and the scene. People are still needed to develop, fine-tune and put them in context, and that takes the fundamentals.
10 articles · 66 min
Start a project
Want this applied to your own project?
Learning is the first step. If you would rather have a working system, book a discovery call and we will scope it with you.