Retrieval-augmented generation: Which technique best fits | AIGP
7-day money-back guarantee — full refund within 7 days of purchase if you've completed under 20% of the questions. See pricing →
Certifications Tools Flashcards Career Paths Exam Guides Blog Pricing For Teams About

Language

✓ EnglishDeutschEspañolFrançaisPortuguês
Check readiness — free →

Retrieval-augmented generation: Which technique best fits this need?

AIGP Understanding How to Govern AI Deployment and Use Medium

RAG grounds answers in retrieved current documents at query time, avoiding retraining when content changes.

The question

An organization wants a general LLM to answer questions grounded in its own internal documents, which change constantly, and it does not want to retrain the model each time a document is updated. Which technique best fits this need?

Preparing for AIGP? Take the free 5-min readiness quiz →

  1. Retrieval-augmented generation, which fetches relevant current documents at query time and grounds the model's answer in them. ✓
    Correct: RAG grounds responses in retrieved up-to-date documents at inference, so changing content needs no model retraining.
  2. Fine-tuning, which further trains the model's weights on the organization's documents so the knowledge is permanently baked into it.
    Plausible but wrong: fine-tuning embeds knowledge in the weights and would require re-running training whenever the documents change.
  3. Edge deployment, which runs the model locally on devices near the data source to cut latency and keep data on-premise.
    Plausible but wrong: edge deployment concerns where the model runs, not how it stays grounded in frequently changing documents.
  4. Agentic architecture, which lets the model plan multi-step tasks and call tools autonomously toward a broader goal.
    Plausible but wrong: agentic design orchestrates actions and tool use but does not itself solve grounding in current internal content.
The trap
Choosing fine-tuning for dynamic knowledge that actually calls for retrieval-based grounding.

How to remember it

RAG grounds answers in retrieved current documents at query time, avoiding retraining when content changes.

How many of these would you get right?

One of 1581 AIGP questions on Certsqill. Take a free five-minute check and see your score per domain — not one number, but which section to open tonight.

Test your AIGP readiness — free

More Understanding How to Govern AI Deployment and Use questions

Part of the Certsqill AIGP question bank · Understanding How to Govern AI Deployment and Use · Every answer, right and wrong, comes with its own explanation.