Privacy Policy

What happens on the site

There is no account, form or email field. The hosting keeps standard server logs, such as IP address, browser type and request time, to operate the site and investigate failures.

On English pages, Yandex Metrica and Webvisor start only after you choose “Allow analytics”. They may record pages viewed, clicks, device and browser information, approximate region, referral source and a replay of interactions with the page. No deck text or Telegram messages are sent to site analytics. You can choose “Only necessary” and the site and bot links will still work.

For attribution, the browser creates a random 12-character site-session identifier. The same identifier is attached to the Metrica visit and to the Telegram /start link, allowing us to tell which consenting site visit reached the bot. It contains no name, phone number, email, Telegram identifier or deck content.

What happens in the Telegram bot

The quick diagnostic happens in @FounderScore_bot. To produce it, the bot processes:

• The text sent after /score. This is a 50–4,000 character startup description used to generate the diagnostic.
• Your bot messages and Telegram identifier. We use them to return the result to the same chat and apply the daily limit.

We do not ask for a name, phone number or email address. Do not include personal or confidential data that is not needed for the diagnostic. A PDF/PPTX for the extended review is processed only after a separate request to @hi_support_team.

The quick diagnostic is generated by an AI language model with no human review. To produce it, the text you submit is sent to a language model provider. We use a mode of operation in which the material sent is not used to train models.

What we never do with your materials

• We do not publish descriptions, decks or reviews, not even as examples. The example on the site is built on our own demo deck.
• We do not pass materials to investors, funds or unrelated third parties. Technical providers process them only to operate the service as described above.
• We do not use the contents of your files to train models.
• We do not read them out of curiosity: a person can open your materials only when you have contacted us about your review, or when we are investigating a technical failure that affected it.

How long we keep it

Your bot materials — the submitted /score text and diagnostic — are kept until you ask us to delete them. The stored site-to-bot attribution link is kept with your bot data and removed by the same deletion flow.

PDF/PPTX material sent separately to support for an extended review is kept only for that service and deleted on request to @hi_support_team.

Technical logs are kept for a limited period and are used only for diagnostics.

Deletion on request

Send /delete to the bot and confirm with /delete_confirm. We delete the stored bot text, diagnostic and attribution link. For material sent separately to support, request deletion from @hi_support_team.

We cannot restore a deleted review, so save anything you need beforehand.

Your rights

• Find out what data of yours we process and why.
• Ask us to correct inaccurate data.
• Ask us to delete your data and materials.
• Complain about our actions to the competent data protection authority.

Bot-data deletion works through /delete; for support material or anything else, message @hi_support_team. We reply on business days.

Changes to this policy

We may update this page — when we change model provider, for example. The version in force is always published here.

How to reach us

All questions about data and deletion: @hi_support_team on Telegram.

Other ways to get in touch are on the contacts page.