2026-09-24
Privacy
What this archive keeps about a person, who else sees it, and how to have it removed.
Who holds this material
The archive is run by the Indian Jewish Heritage Center, together with the Cochin Jewish Heritage Center. The Center decides what the archive contains and what is published.
What is collected, and why
Four kinds of thing, and nothing else:
- What a contributor sends: the file itself, and the description, dates, places and names given with it. This is the archive’s subject matter and it is kept for as long as the archive exists.
- An email address, and a name if one is given. The address is required in order to contribute. It is used for three things only: a Content Expert coming back with a question about the item, a message with the receipt link when the item is sent, and a message with the item’s public link when it is published. It is never published beside a record and it is never used for a newsletter.
- An account, for the Content Experts who review items: a name, an email address, and a password the Center never sees in readable form.
- A record of what failed. When a page or a request fails, the archive keeps the time, the page address without anything after its question mark, the error message, the name of the browser and, for a signed-in account, which account it was. No file and no form contents are kept. Whoever met the failure is shown a short code to quote when they report it. Only administrators can read this record, and it is deleted after ninety days.
Family names
A family name attached to a record is published with it, because a heritage archive that cannot be searched by family is not much of a heritage archive. That also means the surname of living relatives can appear in public.
A name is attached by a Content Expert, from a list the Center maintains, and only where the name is already part of the historical record. A contributor’s own name is not published merely because they gave it.
The machine that reads the file
Every uploaded file is read by an automated service, Google Gemini, which proposes a description, a period, a place and keywords. A person checks that reading against the original before anything is published; nothing the machine produces is published unread.
To be read, the file is sent to Google. It leaves the Center’s systems at that moment.
While reading, the service may also search the web for background on what it sees — a place, a community, a kind of object. What such a search finds is shown to the Content Expert as unverified background, apart from the catalogue, and is never written into a record as fact.
The Center currently uses this service on Google’s free tier. Under the free terms, Google may use what is sent to it — including the file itself — to improve its services. Anyone who would rather their material were not read this way should not upload it, and can write to the Center instead.
Where the data sits
Records, files and accounts are held by Supabase in Frankfurt, in the European Union. The site itself is served by Vercel, also from Frankfurt. Both keep technical logs of requests, which include IP addresses, for their own operation and security.
The messages the archive sends a contributor go out through the Center’s Google (Gmail) account, so the recipient’s address and the message pass through Google.
The Center also keeps its own copies of the database and the files, so that an accident at a supplier does not end the archive.
Cookies
The archive sets as few as it can:
- ijhc.lang — the language chosen for the site. A year.
- ijhc.theme — light or dark, if either was chosen. A year.
- A session cookie, for a signed-in Content Expert only. It is what keeps them signed in.
- If the Center turns on visitor statistics, those cookies are set only after a visitor agrees, and the choice itself is remembered in a cookie. Declining leaves the archive fully usable.
Who else sees it
Content Experts and administrators of the Center see everything in the review screens, including contributors’ email addresses. Nobody outside the Center is given this material, and none of it is sold or exchanged.
A published record is public by definition: anyone can read it, link to it and copy it.
Asking for something to be removed
Anyone may ask what is held about them, ask for it to be corrected, or ask for material they sent to be taken down. Anyone may also ask to be erased without withdrawing the material: the name and address are deleted and the item stays in the archive.
A contributor also holds a receipt link, given when the item was sent. It shows what happened to that submission and needs no account.
Children
The archive is not aimed at children and does not knowingly collect their details. Material showing a child who is now an adult is family history and is treated like the rest of the archive; a request to withdraw it is honoured on the same terms as any other.
Changes
This page carries the date it was last written. A change that affects what happens to material already sent is announced on the site rather than made quietly.
