Why Senso
The trusted AI answer
What Senso is for, and why it exists. Read this before the setup pages.
Why
AI is going to make decisions for us. Those decisions are only as trustworthy as the information behind them.
How
Senso gives AI agents verified truth with proof.
Checked
Claims tested against a source of record. Contradictions surfaced, not silently resolved.
Attested
An accountable owner signs off, whether a person or an agent. The decision is recorded with the claim.
Traceable
Every answer carries the documents behind it, with ids and versions.
A document with all three is a Verified Source. Content that has not been through that gate is not one, however well written it is.
What
A Context Layer your agents query, where every answer traces back to a document you own.
Ingest what you already have. Any agent queries it over MCP or the API. Publish what the world should read.
If you already run retrieval, Senso does not replace it. It evaluates it — the same question against your stack and against your approved sources, and the difference between the two answers is the work.
Where your ground truth lives
“Approved sources” is not a metaphor. It is five specific places, and which one a fact goes into decides what it can be checked against later. A price in a PDF cannot be checked against a catalog, because as far as the system is concerned there is no catalog.
Together they are your Context Layer. Anything a model says that contradicts them is an observation, not a correction.
Why it matters
Today, when a model says something wrong about your company, you cannot produce the document behind it — there is no link between the answer and your Context Layer. Every wrong answer is a fresh investigation.
With that link, a wrong answer is a bug with a stack trace. Find the document, fix it once, and every future answer changes.