Terms of use

1. Who these terms are between

These terms are between [Operator legal name not configured] (“we”), of [Postal address not configured], and you, whether you use FAUVION as a program manager or as a founder.

2. What the service is

FAUVION runs practice investor meetings. The meeting is a live spoken conversation with an AI model playing a generic, sceptical practice investor. It asks questions based on information you supply, follows up on what you say, and replies aloud in a synthetic (AI-generated) voice. You are talking to an AI, not to a person. When the meeting ends, the same model produces one written report on how your answers held up.

A meeting needs a microphone; there is no typed alternative. For the length of the meeting your microphone audio is streamed from your browser directly to the live-voice provider (OpenAI), which also receives your written brief, the text extracted from your pitch deck if you uploaded one, your chosen meeting focus, and a pseudonymous account identifier (a one-way hash, with no name or email) that it uses for abuse monitoring. OpenAI handles all of that under its own terms, which are not part of these and which we do not control. FAUVION itself never receives the audio: the service does not record, upload or store it, and it makes, shows and stores no transcript.

The report covers the content of your pitch, whether each question was addressed, and coaching on how you spoke — pace, pauses, clarity, filler words, structure and concision. That coaching is the model’s impression from listening, not a measurement. The report is written in the model’s own words and contains no direct quotes or excerpts of anything you said or wrote. Separately, your browser measures a few figures from the microphone signal on your device — such as how long your turns and pauses were — and keeps only the numbers, which are stored inside the report.

OpenAI limits a live session to 60 minutes and offers no way to extend or resume one. Within it, the spoken interview has a hard maximum of 55 minutes, and up to 5 minutes are kept for the same model to write the report. At 55 minutes the conversation ends automatically. Each session also has a usage limit covering everything the AI does in it; a session that reaches it is wrapped up earlier, with notice on screen, and a session too close to it cannot open another conversation. A meeting therefore lasts up to 55 minutes and is not guaranteed to run the full time. If the report cannot be finished within the 60-minute limit, there is no report for that meeting; an unfinished report is never provided. You can mute your microphone, or end the meeting, at any time; the end of the meeting produces the report. A meeting you leave before ending it, or whose connection drops, cannot be resumed later, because no recording or transcript of it is kept. You can start the conversation again — a session allows up to 3 live conversations — but each session produces one report.

3. What the service is not

• It is not investment, legal, tax or financial advice, and nothing it produces should be treated as any of those.
• It is not a person. The investor you speak to is an AI model with a synthetic voice.
• It does not represent, simulate or speak for any real investor, investment firm or individual, and it is not affiliated with any. A firm you choose is only a practice focus: it uses a small amount of context from the firm’s own public website where this project holds some, which the firm did not supply, review or approve, and is a general conversation otherwise.
• It makes no prediction, score or assessment of whether you will raise money, and no statement it makes should be read as one.
• Coaching on how you spoke is a practice suggestion, not an assessment of your emotions, personality, confidence, voice, accent or fundability. It rests on what an AI model heard, alongside approximate measurements made in your browser (timing, pauses, pitch movement, loudness variation). The model can mishear, and the coaching can be wrong.
• It does not assess body language or appearance — there is no video — and it creates no voiceprint and does not identify anyone by their voice.
• It does not guarantee any outcome, introduction, meeting or funding.

4. Accounts

You are responsible for what happens under your account and for keeping your credentials to yourself. Founders join a cohort through an invitation link; do not forward one to someone it was not meant for. Program managers are responsible for who they invite.

5. Your content

You keep ownership of everything you upload, write or say. You grant us a limited licence to process it for the sole purpose of running your session and producing your report, including its being sent to the live-voice provider (OpenAI) as described in the privacy notice. That licence ends when the content is deleted.

Your microphone audio reaches OpenAI from your browser and is never held by us. By default OpenAI does not use it to train its models, but it may keep abuse-monitoring logs, which can include content, for up to 30 days under its own terms. We cannot delete what it holds.

You confirm you are entitled to upload what you upload. Do not upload material belonging to someone else who has not agreed to it, and do not upload anything unlawful.

6. Retention and deletion

Session content — your brief, your deck and its extracted text, and your report, including the figures your browser measured — has a 7-day deletion deadline. There is no audio and no transcript to delete, because the service keeps neither. The content is removed by a scheduled cleanup job once that job runs after the deadline, or immediately when you delete it yourself, on the terms set out in the privacy notice — including when the scheduled job runs and its explanation of provider database backups. Read that section before uploading anything you would not want held in a backup for a further short period.

7. Acceptable use

A meeting is a conversation with an AI that speaks in a synthetic voice, not with a person, and it needs a working microphone. Whatever that microphone picks up is streamed to the live-voice provider (OpenAI) while the meeting runs.

You agree not to:

• attempt to obtain another user’s brief, deck or report;
• bring another person’s voice into a meeting without their agreement;
• probe, scan or attack the service or its providers;
• use the service to generate material that harasses, defames or discriminates;
• automate access in a way that degrades it for others.

8. Availability [REVIEW]

The service is provided as-is. We do not commit to any uptime, response time or continuity of the live-voice provider (OpenAI), and we may suspend or change it. A meeting that is cut short by a dropped connection cannot be picked up where it stopped. A real service level commitment, if you want one, belongs here.

9. Liability [REVIEW]

To the extent the law allows, we are not liable for decisions taken on the basis of feedback produced here, for lost opportunity, or for indirect loss. The limits and carve-outs that actually apply depend on your jurisdiction and need drafting; this paragraph is a placeholder for them.

10. Governing law [REVIEW]

Stated as [Jurisdiction not configured]. Confirm this against where the operator is established and where its users are before relying on it.

11. Contact

Questions about these terms go to [Contact email not configured], or through the contact page.