Hello Readers!
My name is Amina, and I’m here to talk about my RAG.
Combining search + LLM = RAG.
First: search for relevant info in a database.
Then: feed it to the LLM → better answer.
Analogy: Instead of memorising everything, the model “opens a book” before answering."
Most developers think of HTTPS as a checkbox. Something you enable because every tutorial tells you to.
Most API breaches happen because systems trust valid requests too much, not because attackers break in. Common wrong…
Seen this before? Let’s set the record straight: Programming languages didn’t evolve to be “better.” They evolved to…
What is Prompt Engineering? Art of writing instructions for LLMs. Bad prompt → bad answer.
Vector DB = a special database to store embeddings (numbers). Let's you quickly find “closest meaning” results.
Embeddings = turning words/sentences into numbers (vectors). These numbers represent meaning, not just letters.
Training = feeding data to a model so it learns patterns. Example: Show a model 1M cat & dog pictures → it learns to…
Inference = using a trained model to get answers. Training = teaching the model.
This is Beautiful👍👍
This is great 👍