Diagram

A sequence diagram with six lifelines — Customer, Support Desk App, Knowledge Retriever, Solution Generator, Gemini API, and Cloud Storage — showing ordered steps where retrieval from Cloud Storage happens before generation by Gemini to deliver a grounded answer.
How Grounding Works in a Customer-Support Application
Retrieval-augmented generation (RAG) is the method managed support applications use to ground answers in a knowledge base. In the Google Cloud example, containerized services run in a Google Kubernetes Engine (GKE) cluster, the support knowledge base lives in Cloud Storage, and the Gemini API performs retrieval and generation. Data flows from the customer to the retriever, then to the generator, and back to the customer, with Cloud Storage as the source of truth on every request. Generation happens only after retrieval, so the final answer depends on enterprise sources rather than on the model alone.
Domain 5 of 5
34 practice questions
Practice this DomainLast updated