Context Engineering vs Prompt Engineering

https://hellopm.co/ai-sprint-resources/?utm_medium=email&utm_campaign=tas2025_resources_rewards

When we talk about prompt engineering, we are only talking about the English instructions that LLM is about to get.

RAG: Retrieval Augment Generation

This is a framework or an approach through which this problem of putting the right thing into context and contextualizing your own database or your own documents.

RAG has two base:

The first part is the knowledge creation.

‘Chunk’ing: to break the whole big document into small parts without losing the meaning.

Vector is a point with a certain length.

We will take a piece of information. We will convert into a vector. And we will take another piece of information, and we will also convert it into a vector. If these 2 pieces of information are related to each other, have closer meaning to each other then in the vector space they would be very closer to each other.

Vectors have many dimensions so not only 2 so we can predic

Second part is handling the user query.

it is converted to vectors to make sure that, eventually when you are finding the right documents. You’ll be able to get ‘semantically’ meaningful documents.

prompts directly convert to embedding (vector)

1- collect all the information

2- convert into chunks

3- convert into vector embeddings with OpenAI APIs

4- It will be stored in vector storage