Artificial IntelligenceRAG
Retrieval-Augmented Generation
Short definition
A technique that augments LLM answers with up-to-date information retrieved from a knowledge base before responding.
Extended definition
In RAG, the user question is converted into a search against a database (like company docs or blog posts), then the results are passed to the LLM along with the original question. The result: accurate, current answers, instead of stale or fabricated (hallucinated) information.
Example & use case
The Ortech support bot uses RAG to search 50 knowledge base articles, then answers the customer with verified information.
Need a custom solution?
Our engineering team can help you apply these concepts to your own product or business.
Talk to us