Glossary
Hallucination
When a language model states something confidently that isn't supported by its sources.
A hallucination is a plausible-sounding but unsupported or incorrect statement produced by a language model. It happens when the model fills gaps from its training data instead of from verified sources. In customer support this is dangerous, because a confident wrong answer can mislead a customer or create liability.
CertifChat mitigates hallucination by grounding answers in retrieved content and instructing the model to answer only from those sources. Inline citations let readers verify each claim, a confidence score flags shaky answers, and when retrieval finds nothing relevant the assistant declines and hands off rather than inventing a response.
See how this works in practice across CertifChat's features.
Related terms
- Retrieval-Augmented Generation (RAG)Grounding an LLM's answers in retrieved source documents instead of its training data alone.
- Semantic SearchSearching by meaning using embeddings, often combined with keyword search.
- Knowledge BaseThe collection of content an assistant draws on to answer questions.
- Ticket DeflectionResolving customer questions automatically so they never become support tickets.
Put your knowledge to work today
Connect your sources, drop one snippet on your site, and start deflecting tickets. We are opening CertifChat to a first group soon.