voice memory
voice memory summarizes the patterns a user accepts or dismisses. it turns repeated editing choices into useful profile signal.
quick answer
use voice memory to explain why the app seems to learn from user choices.
use this for
- reviewing accepted and dismissed patterns
- debugging why suggestions changed over time
- training the profile without manual profile edits
steps
- run checks across real drafts.
- accept fixes that match the user's voice.
- dismiss fixes that are false positives.
- open the voice tab.
- review the voice memory panel after enough signals exist.
details to know
- the panel stays empty until enough signal exists.
- accepted patterns are shown as genuine voice issues.
- dismissed patterns are shown as not your voice.
- voice memory is based on behavior after onboarding too.
limits and edge cases
- one click is not enough to train a rule.
- users can still edit the written profile directly.
ready to use the product instead of reading about it?
open app