Glossary
Grounding.
The practice of anchoring every AI-generated answer to a verifiable source in the underlying data. Instead of the model writing from its training data, every claim it makes traces back to a specific passage, row, or relationship in your actual content.
Without grounding, an AI answer is a guess that sounds confident. With grounding, every claim has a footnote. The difference between an AI that knows and an AI that's bluffing.
What this looks like in practice
What grounding looks like in practice.
Every answer includes citations to the specific source passages it pulled from. Click the citation, see the original text in context. No claim is unsourced. No source is invisible. If you can't trace the answer back to a specific passage, the system didn't actually answer your question. It generated something that sounds like an answer.
The harder problem: even with citations, models confabulate.
Naive grounding gives the model retrieved passages and asks it to write an answer. The model still phrases the response, so it can still invent claims that "feel" cited but aren't actually in the source. Production-grade grounding has to verify the answer against the source AFTER generation, not just before. Most products skip that step.
Why "tells you when it doesn't know" matters.
The most dangerous failure isn't a wrong answer. It's a confident wrong answer when the system should have refused. Grounded systems can detect when the retrieved sources don't actually answer the question. Instead of guessing to seem helpful, the system reports the gap. Lumen tells you what's missing instead of making it up.
Knowing, not guessing.
Grounding is the foundation of trust. Every answer Simply Asking gives cites the source it came from. When the source doesn't exist, the answer doesn't exist either. We tell you what's missing so you can fill the gap.
Bring your knowledge.Simply ask.
CASA Certified · SOC 2 Infrastructure · GDPR-Aligned · Enterprise Ready