FAQ
Questions people actually ask
Including the ones with an annoying answer. A page that only answers the easy questions is not an answer.
- Which apps does she work in?
iMessage and Telegram today. You text her the way you text anyone else — no separate app, no dashboard, no new number to learn.
WhatsApp is coming soon. It is not switched on, so we do not count it as working; when it is, it will say so here and on the features page.
- Can she send email for me?
No. Her connection to your mail is read-only, and that is enforced by the permissions you approve when you connect it — not by a setting we promise to respect. A request to send, delete, label, or change anything fails at Google's end.
So she reads and summarizes, and that is where it stops. If you want an email sent, she can draft it and you send it.
- Can she see my calendar?
Yes, if you connect it — and reading is all the connection carries. Ask her what is on tomorrow and she reads the calendar you connected back to you, the same way she reads your mail.
The permission you approve covers reading, so an appointment is something she can tell you about rather than something she can move. If you would rather she did not have it at all, untick calendar on the consent screen and she works without it.
- Can I get a daily brief of my inbox?
Ask her for one. Today the brief is on request: you text her "what's in my inbox" and she reads it back to you — what needs you, what can wait — in whichever language you asked in.
A standing brief that arrives on its own every morning is not something we set up for you yet, so we are not going to say it does.
- What languages does she speak?
Your preferred language — Farsi, Spanish, English and the other major languages spoken in Los Angeles. In text and in voice: send her a voice note and the reply comes back as a voice note, in the language you used. Switch mid-conversation and she switches with you.
This is not a translated interface with an English assistant behind it. She listens, thinks, and answers in the language you used.
- How do I get access?
Pick the plan that fits on the plans page. Starter, Pro and Max are yours to choose; the Business and Enterprise plans start with a short conversation, because what they include depends on how your business shares its records and where your data has to live.
Setup is personal either way. We connect her to your accounts together, and you decide what she is allowed to see before anything is switched on.
- Why does Gmail ask me to re-approve access every week?
Because of a Google policy, and it is genuinely annoying, so here it is upfront rather than as a surprise in week two.
For a personal Gmail address, the connection Rumi uses sits in Google's testing state, and Google expires that permission after seven days. The alternative Google offers is a paid annual security audit of the whole project, which costs far more than it is worth for one inbox — so instead you tap "allow" again about once a week.
If your mail is on a Google Workspace account with your own domain, this does not happen: that connection lasts, and you approve it once.
- What happens to my data?
Today Rumi runs as a service we operate for you, on machines we administer. What we can give you is a commitment rather than a property of the setup: a small team runs it, we do not read your messages, that is written into your agreement, and we delete your data when you ask.
The version where it is a property — Rumi running in your own cloud account, where we hold no standing access — is on the roadmap and is a different deployment, not a switch we can flip on the shared one.
The security page says all of this in more detail, including the parts we cannot promise.
Want her in your chat?
Two plans you pick yourself, and a third we shape around your team. Setup is done with you rather than handed to you.
Recruit your assistant →One line per person, billed monthly. Farsi, Spanish, English and more — whichever you speak.