Fyxer AI gets better the more you use it. While it’s designed to work out of the box, there are a few ways you can actively teach it to improve - especially around categorizing your emails, drafting replies, and accurate meeting notes.
Email Categorization: How to Guide Fyxer
Fyxer uses smart rules to sort your inbox, but it doesn’t always know what you consider high priority. Luckily, it learns fast.
Ways to train it:
Create your own rules under Preferences → Email Rules (by sender, subject, or domain).
Drafts: Help Fyxer Write Like You
Fyxer learns your tone by observing how you edit its drafts.
Tips for faster learning:
Tweak, don’t delete. Instead of deleting a draft and starting fresh, make light edits - add context, reword, or shorten. Doing this consistently helps Fyxer adapt.
Batch review. Reply to at least 5 drafts at once - this gives Fyxer a strong sample.
Use the “Draft Prompt” box to guide tone, intent or format.
Better with Meeting Notetaker on - Context collected from meetings improves the relevance and quality of draft replies
Meeting Notetaker: Improving Accuracy
Fyxer’s Notetaker captures meetings, summarizes key points, and drafts follow-ups, but you can make it even sharper.
Best practices:
Add custom terminology in Meetings → Settings. Product names, acronyms, internal shorthand - this all helps.
Avoid very short calls. Fyxer works best with at least 10 minutes of context.
Let the Notetaker in on time to ensure nothing’s missed.
Memory & Privacy: What Fyxer Remembers (and Doesn’t)
Fyxer remembers what helps - and forgets what doesn’t.
Only you see your edits. Your personal memory stays personal.
No shared models. Fyxer doesn’t use your data to train external tools or public AIs.
Fyxer never uses data from deleted accounts or restricted content. You’re in charge.