You are deciding whether to put case evidence into someone else's system. These are the answers to the questions that decision actually turns on, stated plainly and without hedging.
Does zimni train AI models on my footage or case data?
No. Customer footage, transcripts, and case data are never used to train or fine-tune any model, ours or a third party’s. Your evidence is processed to answer your questions and for nothing else.
Does my video get sent to an outside AI company?
No. Video and audio are processed inside zimni infrastructure and never leave it. Only the text transcript derived from your footage is sent to a language model provider. The footage itself stays with us.
Which AI providers does zimni use?
Text processing runs on Anthropic, Google, or OpenAI, selected at our discretion so we can route to the best model for a given task. Whichever is active, we contract for API terms under which your data is not used for model training.
Where and how is my footage stored?
Your evidence is stored on encrypted servers protected by SSL/TLS, both in transit and at rest. We are finalizing our hosting region and will publish the exact details here once confirmed.
What happens when I delete footage or close my account?
Deletion removes the material from primary storage and from backups. It is not retained after deletion in either location.
Can zimni employees watch my footage?
Not in the ordinary course. Staff access requires an explicit break-glass escalation, and every such access is recorded in an audit log.
Is zimni SOC 2 or ISO 27001 certified?
Not yet. zimni is an early-stage company and does not hold formal third-party certifications. Our engineering and security practices are built to meet those standards as we scale, and we would rather say so plainly than imply otherwise.
Questions this page doesn't answer? Ask us directly — we would rather answer a hard question than have you assume.
See also our general security practices.