before labelr
your inbox is a wall of unread emails. newsletters mixed with receipts mixed with actual important stuff. you either spend time sorting manually or just let it pile up.
after labelr
every new email gets automatically classified and labeled. newsletters go to newsletters, receipts go to receipts, work stuff gets flagged. it runs in the background, you don't think about it.
how it stays private
the whole thing runs on your machine. no email data leaves your computer. if you use ollama, even the AI inference is local. you can also plug in groq, deepseek, openai, or any openai-compatible API if you prefer speed over privacy.
setup takes 2 minutes
run labelr init and a setup wizard walks you through gmail oauth, picking your AI provider, and defining your labels. after that, labelr start installs it as a background service. launchd on mac, systemd on linux, task scheduler on windows.
retroactive labeling
already have 10,000 unsorted emails? labelr sync --since 6m goes back through the last 6 months and labels everything. works with any time range.