TPEE Toptronic® Prompt Engineering Editor

8. RAG, Hallucination, and Verification

Reduce unsupported answers and verify important work.

By Jacques Botte, founder of Toptronic®. Last updated 12 September 2026.

The lesson

Hallucination means the AI produces plausible text that is not supported by the facts. RAG reduces this by giving the model documents to quote or use.

For business, legal, electrical, medical, or production code decisions, always ask for sources, inspect the source, and test the result.

TPEE's Data and Examples sections are where you tell the AI which facts matter most.

Check yourself

Question 1: What does hallucination mean in AI work?
  1. A saved preset
  2. A valid citation
  3. A plausible answer not supported by facts — correct
  4. A faster GPU

Answer: A plausible answer not supported by facts

Hallucinations sound convincing but may be unsupported or wrong.

Question 2: How does RAG reduce guessing?
  1. By forcing offline mode only
  2. By supplying relevant retrieved documents as grounding context — correct
  3. By removing all context
  4. By changing the font

Answer: By supplying relevant retrieved documents as grounding context

Retrieved facts can anchor the generated answer.

Question 3: What should a professional do with cited sources?
  1. Inspect them before trusting the answer — correct
  2. Assume they are always real
  3. Delete them
  4. Ignore them if the text sounds confident

Answer: Inspect them before trusting the answer

Grounding only helps if sources are relevant and verified.

← Previous lesson · All 83 lessons · Next lesson →

The full course — 83 lessons and 249 quiz questions — ships inside the app. Get TPEE to study it offline.