Not used to train AI models
We do not train models on customer data, and our agreements say so.
Trust
Built for firms that answer to regulators.
We don't train AI models on your records, every run leaves a trail your compliance team can follow, and anything that doesn't fit is flagged for a person to review.
SOC 2 Type II
Independently audited. Request the report, our policies and the current subprocessor list through our Trust Center.
Data and access
What happens to your data.
Stored in the United States
Customer records are stored in the United States.
Deleted when the agreement ends
At the end of our agreement, your data is deleted and we confirm it in writing.
Controlled access
Sign in through your own identity provider, with its multi-factor rules, and role-based access.
A full audit trail
Every run records where each figure came from and who approved what.
Approval where you want it
Anything that writes back to your systems can wait for a person to approve it.
Exceptions, never silence
If it doesn't fit, you hear about it.
A step that falls outside what your team approved is held with a reason and an owner, so gaps surface before a client document goes out.
New account opened 9/17. Outside what you approved. Held for your review.
Approved by the lead advisor at 08:42. Added to the packet and logged.
Bring your hardest weekly workflow.
See how it runs with an answer key, on a call with the people who build it.