Zero-Knowledge Marketing Tool for Regulated Industries
In law firms, FinTechs or medicine, the first question of any tool is: where do the data end up? YourValidator answers it clearly - nowhere but in your own browser. The modeling runs 100% client-side, with no server transfer and no cloud requirement. So you can calculate even where data must not leave the building.
Why the cloud solution is weaker in regulated industries
Where data sits is a compliance question
A tool that uploads inputs to a server shifts the responsibility to a third party - and raises exactly the question that stalls every sign-off in regulated industries: who has access, where is it stored, how is it deleted? Clarifying it alone often costs more than the calculating is worth.
What 100% client-side changes
If the data stay in the browser, the question falls away. There is no server where anything could sit, no transfer to be secured. The responsibility stays with the user - and thus where the regulation locates it anyway.
How the local modeling works
- Open it in the browser. No sign-in, no email - the calculator starts right away.
- Compute locally. All inputs stay in localStorage; there is no server transfer. After the first load, the calculation also works offline.
- Save encrypted. The backup can be saved with a password as an AES-256-encrypted .validator file - or without a password as a readable JSON file.
- Exchange locally. Pass scenarios along as a file instead of sharing them via a cloud - control stays with the sender.
- Use the AI only deliberately. The optional AI function is the only point where data leave the device - and only if you actively decide, in the prompt field, to call your provider.
Model locally in the browser, without data leaving the device: try the tool
Prerequisites & limits
What it needs
- A current browser - nothing more. No server, no login, no cloud account.
- An active backup: localStorage is not an archive. Whoever clears the browser cache without saving first loses the inputs.
- For sensitive states, the password-protected, AES-256-encrypted .validator format instead of the open JSON file.
Where the limit lies
- The optional AI function deliberately breaks the local boundary: if you use it, your prompt goes directly to the chosen provider (Gemini, OpenAI or Anthropic) - via your own API key, not through our servers. You make that decision actively in the prompt field.
- The API key is stored solely locally and never leaves the device. Without using the AI, the modeling stays fully client-side.
Common thinking traps
- Mistaking localStorage for an archive. It holds the last state in the browser, not permanently. Without a backup, everything is gone as soon as the cache is cleared.
- Mistaking the open JSON file for protected. Without a password, the backup is readable plain text. For sensitive data, the encrypted .validator format belongs used.
- Extending zero-knowledge to the AI. The local guarantee applies to the tool itself. As soon as the optional AI runs, the prompt leaves the device toward the provider - that's no contradiction, but it belongs known.
Frequently asked questions about zero-knowledge
Are my inputs transmitted to a server?
No. The modeling runs entirely in the browser; all inputs stay in localStorage, there is no server transfer. After the first load, the calculation also works offline.
How do I save sensitive scenarios?
Via the backup with a password: the file is saved AES-256-encrypted in the .validator format. Without a password, a readable JSON file results - for sensitive states, the encrypted variant is the right one.
Do data leave the device when I use the AI function?
Yes, but only then and only on your active decision. If you use the AI, the prompt goes directly to your provider via your own API key - not through our servers. Without the AI, everything stays local.
Where is my API key stored?
Solely locally in the browser. The key never leaves the device and is never transmitted to our servers at any point; we have no access to it.
Is this suitable for regulated industries?
That's for your own compliance to decide - the model creates the precondition by running locally without a cloud requirement. The responsibility for use and legal assessment stays with the user.
