Questions, answered.

A workspace with a helper already inside it. You press the microphone and say what you need, or type it, and the helper writes your notes, plans your week, drafts your invoices and keeps your calendar and contacts in order. It is built on Geenius OS, in a plainer and more playful voice.

For people who have never used an AI, and for anyone who would rather talk than type. A plumber in a van, a salon owner, a club treasurer, a parent with a busy week. You need no AI account and no prompts.

Request access. You join the waitlist and we invite people step by step. There is no free plan.

AVS Media, in Nijmegen, the Netherlands. Pilotie is built on Geenius OS, which is made for the EU and the GDPR. See your data and the privacy statement.

There are short drawn videos on See it done, with made-up examples. A live demo of the helper is planned, with no date.

It is the same workspace. Geenius OS has no AI inside and you bring your own, for €19. Pilotie adds the built-in helper and the microphone for €29. See Pilotie or Geenius OS?.

Yes. The helper is built in and is part of the plan: you do not open or pay for an AI account on the side. It runs on Claude models: a fast small one for everyday jobs, a bigger one for long writing, and the biggest only when you ask for the best.

It reads your workspace and uses its tools: notes, tasks, lists, reminders, calendar, contacts, documents, invoices, forms, files and the apps you installed. It keeps its answers short and tells you what it did. See the helper inside.

It is told never to invent data and to say so when a tool returns nothing. It can still be wrong, as any AI can, so read drafts before they go out. That is why sending always waits for your tap.

While a workspace is in test mode the helper gives clearly marked practice replies instead of real AI answers, and says "Test mode" in each one. Everything around it still works, so people can try the flow. Early-access workspaces may start in it.

Claude models. A small, fast Haiku-class model does the everyday jobs, a Sonnet-class model does long writing, and an Opus-class model only when you ask for the best. See the helper inside.

The empty chat offers three starters that fit your kind of workspace. A good first one is "What is on my plate today?". See what to say.

Hold the microphone and talk, or tap to lock it open. The recording is written out as text by a speech service, shown to you, and the helper answers it like any other message. See voice messages.

The app is set not to keep the audio after it has been read. What stays is the text of your message, like a typed one.

Up to 2 minutes 10 seconds and 6 MB. A workspace can send up to 150 voice messages a day.

Your recording goes to a speech service to be written out: by default Groq (Whisper large-v3-turbo), with OpenAI (gpt-4o-mini-transcribe) as the fallback. The names in your workspace go along as a spelling hint. See the privacy statement.

Not yet. Live dictation and a spoken reply are planned, with no date.

No. Anything that sends, pays, publishes or deletes for good becomes a card, and you tap Yes or No. Reading and drafting do not need a tap. See you stay in charge.

It waits for a day and then expires. Nothing is sent.

No. Only a person answers.

Text that comes from mail, web pages or forms is treated as data. The helper tells you what it asked for and lets you decide.

Yes. Every step is in the history, with who did it: you, the helper or an AI you connected. See audit trail.

A few short facts, such as "I work Tuesdays to Fridays". They are listed on the page What I remember, where you can change or forget any of them and add your own. See memory.

Open What I remember, press Forget next to the fact and confirm. It asks first.

The helper answers in the language you wrote in and its screens come in English, Deutsch and Nederlands. Speech writing knows many languages, but we test and promise only those three. See your language.

Not yet. This website is in English for now.

€29 per seat per month, VAT included for consumers. That includes the workspace, the helper and 600 credits a month. See pricing.

No. There is also no trial that we promise on this site.

Half a cent of what the helper costs to run. Every message costs at least one credit, and the card shows what today used. See credits.

The helper takes a break and tells you how to carry on: a pack of credits, a bigger plan or waiting for renewal. Before that, from 80 percent used, you get a gentle note.

Packs of 1,000 and 5,000 credits exist in the catalogue, but the prices and the checkout are still to be confirmed, so we do not quote them yet.

Yes. Pilotie is a hybrid: the workspace stays open to any AI that supports MCP. Copy the address from the page Use your own AI too and add it as a connector. Both work at the same time. See use your own AI too.

The Model Context Protocol: the open standard AI assistants use to connect to other software. You never type any of it.

No. Your helper keeps working. The history shows which of them did what.

In your workspace, not in the helper and not in a chat. You can export them.

It drafts them as PDFs with the legal lines for Germany and the Netherlands. The email waits for your tap. Templates are not legal advice.

Yes. It checks the new time first, blocks double bookings and drafts the message to the person. Undo is one click.

No. It reads your contacts and drafts messages. It does not call, and it does not send without your tap.

Yes, with a preview first. Nothing is public until you say publish. A new shop is a test shop on a mock payment until a person switches it live.

No. Designs are yours to use. Nothing is posted for you.

Your message and the parts of your workspace the helper reads to answer it go to the model provider, as with any AI. Your voice goes to the speech service. The privacy statement says what is stored.

We do not claim anything about training on this site yet. The privacy statement is a draft, and what we say about training will be added once it is confirmed.