What does retrieval mean in a RAG pipeline?
PICTURE THIS: 1, 2, 2, 8
Simple meaning
Retrieval is the step that selects the most relevant chunks for a query.
WHY — Tokens instead of words?
LLMs use tokens (not full words) because it helps them:
who only read docs from people who shipped.
and tied to GenAI / LLM work.
Each piece maps to a number the network can learn.
Fixed pieces are what transformers expect as input.
STEPS — What happens step by step?
Before you speak the answer, walk the interviewer through these steps:
- 1Retrieval is the step
that selects the most relevant chunks for a query.
- 2Typical methods are vector
search, keyword search, or both.
- 3Those chunks become context
for the generator.
- 4Give an example
One tiny concrete case you can say aloud.
- 5Common mistake
What juniors usually get wrong.
- 6Close
When you pick this over the alternative.
EXAMPLE — See it in action
Here's a short line you can speak, broken into clear beats:
Note: Adapt this scaffold to your own project — keep it under 60–90 seconds.
Key takeaway
Retrieval is the step that selects the most relevant chunks for a query. Typical methods are vector search, keyword search, or both.