Use Cases
Common use cases
What a Senso context layer is for today. Two workflows, both of which run end to end.
Available today
The second builds on the first: you cannot check an answer against your facts until your facts are somewhere an agent can read them. Start at the top.
Ask your agent
Show me the questions my industry is asked where we are not cited, and who got cited instead.
How a result is labeled
Every answer shown in a use case carries the mode it came from — on the card, in the export, and on any link you share.
Observed answer
What a surface actually returned. An answer you captured yourself stays labeled as imported.
Controlled model test
A provider API called with a recorded configuration. Not evidence about the consumer product built on it.
Connected-agent test
A real assistant using your sources over an authorized connection. Says nothing about discovery or ranking.
Sample
A scripted walkthrough over fictional data. No model was called, and it never sits beside a live result.
