FAQ
Questions, answered plainly.
What Jesmond does, what it deliberately will not do, and how the confidentiality architecture holds up under questioning.
The detail behind Jesmond
How can I trust Jesmond won’t hallucinate?
Jesmond is not a chatbot. It does not search the internet and does not offer answers of its own, the two habits that produce most hallucination. It uses only material in the sources you have uploaded. Anything it cannot find there is marked as missing rather than guessed. Every document is a draft you check before use.
Does Jesmond make the decisions?
No. Your managers make the findings and decide the outcome. Jesmond supplies the questions and the structure, and checks for consistency.
What happens if Jesmond gets something wrong?
Every document shows its working and identifies the material it drew from, and your decision maker approves it before use.
Is what Jesmond produces disclosable?
Yes, exactly as your files are now. What changes is the quality of what you disclose: a structured record of the questions asked, the evidence considered and the findings made, attributed to a named decision maker.
Can the team at Jesmond.AI (Defend This) Ltd see our case content?
No. It is encrypted on your device under a key derived from your passkey, and we never hold that key. Our servers hold identity, case metadata and ciphertext, so there is nothing for us to read. There is no admin override and no support back door.
What does implementation involve?
We configure your policies, definitions and escalation rules before your first case, and can match your house style from anonymised past decisions. Training for HR and line managers is included.
Still deciding?
Join the priority list and we will answer more in half an hour than a page of questions can.