Minerva
how it works

Say it once. Minerva does the rest.

You don't have a good way to capture your thoughts and turn them into action — especially when you're moving too fast to stop and write or type. Minerva is the missing step: you talk, and what you said becomes tasks, events, and notes in the apps you already use. Here's the whole thing, start to finish.

  1. Talk. Or ramble.

    Record a voice memo the way you naturally speak — one passing thought or a wandering brain dump, from your watch, your phone, or your computer. No commands, no special phrasing, no forms. Mid-sentence course corrections and "um"s are fine; Minerva is built for how people actually talk.

  2. Minerva sorts it out.

    task event note

    Minerva transcribes the recording, then actually reads it. One rambling memo can hold a to-do, an appointment, and a stray idea — each one is pulled out and treated as its own item. Tasks and events get written up as clean summaries, the kind that read well on a list. Notes are different: they keep your own first-person words, because your thinking shouldn't be flattened into someone else's summary.

  3. Filed where you already look.

    Each item lands in your own apps: tasks to Google Tasks, events to Google Calendar, notes to Notion. Nothing waits for you to find time to type it up — in live use, it's about half a minute from tapping "stop" to the item sitting on your list.

when it's unsure

Questions, not guesses.

When something in a memo is ambiguous — a day with no time, a name Minerva can't place — it doesn't silently assume. It sends you one plain-language question, as a notification you can answer whenever suits you, on whichever device is closest: iPhone, Watch, or Mac. Your answer syncs everywhere and is applied to the item within seconds.

And it only asks when it genuinely needs to. When you do say a time, Minerva reads it the way a person would — "dinner at 5" lands at 5 PM, not 5 AM. But it never invents a time you didn't say.

made yours

It learns how you talk.

Everyone means something different by "morning." So when you start, Minerva runs a one-time "getting to know you" phase — about twenty short questions, answered on your own time from any of your devices. It learns your personal phrasing and filing style: where your kind of item belongs, whether a spoken time should be locked in, and what you mean by the words you use. None of it ever holds up a real item.

questions

Common questions

That's normal — a single rambling memo can produce a task, an event, and a note, each filed separately.
No. Talk the way you naturally would; Minerva sorts it out.
Processing lands within seconds of upload — about half a minute from "stop" to filed.
No — that's the point. It feeds the ones you already use.
All questions
think out loud

Your voice should be enough

Join the waitlist for the public release, and we'll let you know the moment Minerva is available.

Join the Waitlist