Dishaya / Trust

Your Work Stays Yours. Here Is Exactly What That Means.

Trust should not require reading between the lines. This page states, in plain language, what stays on your device, what we store and why, how we protect it, and what we will never do. If any of this changes, this page changes first.

What we store, how we protect it, and what we refuse to do.

The Short Version

Your chats and documents stay on your device. Research reports are stored with your account for one reason only: so your Library can reopen them. We never train on your content, in any form. And everything is deletable, by you, whenever you want.

What Stays On Your Device And What We Store

Dishaya is built local-first. The honest way to explain that is two short lists, and this is the whole of both.

On Your Device

With Your Account

That is the full inventory. If something ever needs to join the second list, we will say so here in plain language before it does.

Security Posture

For the small amount we do store, the posture is conservative and standard, and we would rather describe it plainly than impressively.

The No-Training Commitment

We do not train on your content. Never, in any form: not your chats, not your documents, not your reports, not anonymized fragments of any of them, not under consent buried in fine print. Your work is never used to improve any model, ours or anyone else's.

This is a promise and a policy. It is how we have operated from the first day, and it is written into our terms so it binds us, not just describes us.

What Dishaya Does Not Do

Where Humans Should Verify

Dishaya checks every claim against its exact source passage before it is written, and what fails the check is disclosed, never hidden. That makes the work more rigorous. It does not make it infallible, and we would rather you know the difference.

Verification is a rigor tool, not an oracle. A claim can be faithfully supported by its cited source while the source itself is wrong, outdated, or biased. The labels, verified, partial, unverified, and contradicted, tell you how well each claim is supported and exactly where to look harder. They do not replace your judgment.

For high-stakes decisions in law, medicine, or finance, treat a Dishaya report as a strong starting point and have a qualified professional review it before you act. Our published methodology states exactly what each label promises and where its limits are.

Reporting A Concern

If you find a security issue, a privacy gap, or anything on this page that does not match your experience of the product, write to [email protected]. Dishaya is small and independent, so your message reaches the people who build the product. We read everything, and we fix what is real.

Common Questions

Does Dishaya train AI models on my content?

No. Never, in any form, for any purpose, under any label. Your chats, documents, and reports are never used to train or improve any model, ours or anyone else's. This is both a promise and a written policy in our terms.

Can I delete everything Dishaya has about me?

Yes. Chats, documents, decks, and notes live on your device, where you can delete them directly. Research reports, your email address, and usage counters are stored with your account and are deleted when you delete them or your account. Nothing is silently retained.

Who can see my research reports?

You. Reports are stored with your account for one reason: so your Library can reopen them on any device you sign into. They are private to your account, used for nothing except serving them back to you, and deleted when you delete them.

See It Show Its Work

Start Free