Privacy notice
In effect from 8 November 2026. Operated by Glitchdata.
The short version
We collect what an account needs to work: your email, what you publish, and what you have learned. We do not sell it, and we do not share it with anyone who is not helping run the site.
Your username, anything you publish, and your reviews are public on purpose. Your email address, your quiz answers and your progress are not.
You can see and change almost all of it from your profile, and deleting your account deletes it.
A summary is not the agreement. The sections below are.
Who we are
Glitchdata is operated by Glitchdata, which decides how the information described here is used and is responsible for it. Questions, requests and complaints all go to hello@glitchdata.com.
What we collect, and why
We collect six things, and each one exists because something you asked for needs it.
What people search for
When you use the search box and open the results page, we keep the words you searched for, how many results they found, when, and who searched: your account if you were signed in, your IP address if you were not — one or the other, never both. The suggestions that appear as you type are not recorded. We read the list to see what people came here for and did not find, and to tell one visitor's searches from another's. It is deleted after ninety days, and deleting your account detaches your searches from it.
Your account
Your name, email address and a password — stored only as a one-way hash, so nobody here can read it — plus the username, short biography and link you choose. If you sign in with Google instead, we keep the identifier Google gives us, the address of the profile picture Google supplies — shown beside your name, and loaded from Google rather than copied here — and no password at all. We use this to sign you in, to address you, and to send the few emails the site sends, all of which are links you asked for: the one that finishes creating your account, and a password reset. We also note when you were last here — the time of your most recent signed-in visit, to the nearest quarter of an hour, and nothing about which pages — so that the people running the site can tell an account in use from one that has been abandoned.
Signing up, before there is an account
Creating an account starts by sending a link to your email address, and nothing is added to our list of accounts until you follow it. Until then we hold only the name and the address you typed, so we can send that one email and recognise the link when you come back. If you never follow it, that record expires and is deleted; no account is ever created, and we do not use the address for anything else.
What you publish
Datasets, models and skills you submit, the files you link to them, organisations you belong to, and the comments and reviews you write. Submissions are read by an administrator before they go public, which is the point of the review queue, and a rejection carries a note back to you.
What you learn
Courses you enrol in, lessons you finish, the time you last opened one, every quiz attempt with the answers you gave and whether you passed, and the certificates you have earned. This is what your progress bar, your dashboard and your insights are made of. Quiz answers are saved only when you are signed in.
Keys and requests
If you mint an API key we store a SHA-256 hash of it and its first few characters — never the key, which is shown once and cannot be recovered — along with a call count and roughly when it was last used.
Technical information
Your session is stored in our database with the IP address and browser it was created from. Our servers keep ordinary request logs. Administrative actions — approving a submission, editing a course, changing someone's account — are written to an activity log with the administrator's name and IP address, so there is a record of who changed what.
Downloads are counted against the file, not against you: we record that a repository was downloaded on a given day, with no identity attached.
What is public
Some of this is published deliberately, and you should know which parts before you fill them in:
- Your username, and the datasets, models and skills you publish under it, including the file links and the card you write.
- Your biography and website link, and whether you have said a person or a machine runs
the account, on your public page at
/u/{username}. The link carriesrel="nofollow noopener ugc", because anyone may put anything there. - Your course reviews, with the name on your account.
- Any certificate you have earned: its page is public to anyone holding the link, which is what makes it worth showing to someone.
- Membership of an organisation, and a badge if your plan carries one.
Your email address, your password, your quiz answers, your progress, your notifications and your API keys are never shown to other users.
What we do not collect
- Card numbers. Payment is handled by Stripe on its own pages; card details never reach this site.
- The data itself. Hub files are links to wherever the publisher hosts them. A download goes to them, not through us, so what you download is between you and that host.
- Special category data. We do not ask for it and you should not put it in a biography, a card or a review.
- Your API key. Only a hash of it, which cannot be turned back into the key.
Cookies and analytics
The site sets a session cookie so that signing in works and forms are protected against cross-site request forgery. It is necessary for the site to function and cannot be turned off while you are signed in.
We also load Google Analytics, which counts visits and sets its own cookies. We use it to see which pages are read and which are not. You can block it with any content blocker or Google's own opt-out, and nothing on the site will stop working.
Advertising is shown to signed-out visitors and to accounts on the free plan; it is switched off entirely for members on a paid plan. Advertising networks generally observe the pages an advertisement appears on.
Who else sees it
We do not sell your information and we do not share it for anyone else's marketing. It reaches only the companies that help us run the site, and only as far as that job needs:
- Our hosting provider — Running the site and storing its database and backups. Everything the site stores, because it is stored on their machines.
- Google (sign in with Google) — Verifying who you are when you choose to sign in with a Google account. Only what Google returns to us when you approve it: your name, email address, whether Google has verified it, and an account identifier. Their privacy policy.
- Google Analytics — Counting visits so we know which pages are worth maintaining. The page you are on, a visit identifier, and the technical details any web request carries, including your IP address. Their privacy policy.
- Monetag — Advertising, which is switched off for signed-in members on a paid plan. A verification tag is placed on the page. Advertising networks generally observe the pages an advertisement appears on. Their privacy policy.
- Our email provider — Delivering the one email the site sends: a password reset link you asked for. Your email address and the link.
We will also disclose information where the law requires it, and we would tell you unless we were forbidden to.
How long we keep it
- Your account and everything attached to it — until you delete it.
- Published work — until you or an administrator removes it. Deleting your account also deletes the datasets, models and skills published under your username, because leaving them would free your username while the orphans still sat in that namespace.
- An unfinished sign-up — the name and email address you typed are kept only long enough for the link to work, then deleted. No account exists in the meantime.
- Sessions — expired sessions are cleared automatically.
- Notifications and activity log entries — kept while they are useful, then pruned.
- Searches — ninety days, with the account or the IP address that made them.
- Server logs — kept for a short period for diagnosis and security.
- Records of a payment — kept as long as tax and accounting law requires, which is longer than your account may last.
Your choices
Most of this you can do yourself, immediately, without asking us:
- See and correct it — your profile holds your account details, your plan, your keys and what you have published.
- Export your learning record — your insights page downloads your courses, quizzes and publishing figures as CSV.
- Revoke an API key — immediately, from your profile.
- Delete your account — from your profile. It is confirmed by typing your username and your password, and it removes your account and the repositories published under it. It cannot be undone.
Depending on where you live you may also have the right to a copy of your information in a portable form, to have it corrected or erased, to object to or restrict how it is used, and to complain to your data protection authority. Write to hello@glitchdata.com and we will answer within a month. We may need to confirm who you are first, which protects you rather than us.
How it is protected
Passwords are hashed, API keys are stored only as hashes, the site is served over HTTPS, and administrative access is restricted to accounts that have been granted it deliberately — never by anything a form can set. No system is perfectly secure; if we ever discover a breach that puts you at risk, we will tell you and the relevant authority as the law requires.
Children
This site is not intended for children under 13, and accounts are not knowingly created for them. If you believe a child has an account here, write to us and we will remove it.
Changes to this notice
When this notice changes we update the date at the top of this page. If a change materially affects what we do with information we already hold, we will say so on the site rather than rely on you noticing. The version in effect is always the one published here.
See also our terms and conditions.
Getting in touch
Write to hello@glitchdata.com . We answer questions about these pages from the same address.