41AI & promptingTry it now

RAG

Have the AI look up relevant material first, then answer from it — fewer hallucinations, more accuracy.

For example

My app uses RAG to retrieve dictionary material, then has the AI write the explanation.

Resources

Copy prompt

Design a RAG pipeline for my docs: chunking, embedding, retrieval, and prompt assembly.

Related terms

The instruction you give the AI; it decides the quality of the output.

For example

Build me a word-card component in React with the word, pronunciation and a save button.

The relevant information the AI needs to finish the task.

For example

Product goal + current code + requirements doc — all of that is context.

Vector database

A database built for storing and retrieving vector data, common in AI apps such as semantic search.

For example

I store each word's vector in a vector database so I can do semantic search later.