The short version
The free checker never sends your datasets anywhere. If you buy the toolkit we keep your email address, a Stripe reference and a license key, so you can download the file again.
The free browser checker
Files you drop into the checker are read by JavaScript running in your own browser tab, inside a Web Worker. Parsing, comparison and report generation all happen on your device. There is no upload endpoint in this application for dataset files, and the worker makes no network requests. Closing the tab discards everything; we never had it.
You can verify this yourself: open your browser’s network panel, run a check, and watch. After the page and its scripts load, no request carries your data.
What we do collect
- A count of page requests to the home page and the Team page, recorded on our server when the page is requested: the path and the referring site, nothing else. Requests whose user-agent string identifies a crawler are counted separately, and for those we keep the user-agent string so the filter can be checked; the classification is a heuristic and can be wrong in both directions. No IP address, no cookie, no device fingerprint.
- Anonymous page analytics via Vercel Analytics: page path, referrer, country, device type. No cookies are set for this and no cross-site profile is built.
- Purchase records, if you buy the toolkit: your email address, the Stripe customer, checkout session and payment intent identifiers, the amount and currency, the license key we issue, and the time of purchase.
- Account records, if you create an account: your email address, and — if you sign in with GitHub — the name, avatar URL and account identifier GitHub returns.
- Download records: the time of each toolkit download and a one-way hash of the requesting IP address, salted with a server-side secret. We keep this to spot link sharing at scale. The raw address is not stored.
- Server logs and product events: coarse events such as “a checkout started” or “a purchase was recorded”, with identifiers but not with your dataset contents.
What we never collect
- Your training or evaluation data, in any form, including hashes or excerpts.
- Card numbers. Payment details are entered on Stripe’s own checkout page and never touch our servers.
- Advertising or cross-site tracking identifiers.
Who we share with
Only the processors it takes to run this:
- Stripe — payment processing. Stripe receives your email and payment details directly.
- Vercel — hosting and anonymous analytics.
- Neon — the Postgres database holding purchase and account records.
- Resend — sending sign-in links and receipts, when email is enabled.
- GitHub — only if you choose to sign in with GitHub.
We do not sell data, and we do not share it for advertising.
How long we keep it
Purchase records are kept for as long as we sell the product, plus whatever tax law requires — they are also your proof of purchase. Download records are kept for 12 months. Analytics are aggregated and retained by Vercel under their own policy. Delete your account and we remove the account record; the purchase record is retained, with the email address it needs to remain useful as a receipt.
Your choices
Use the contact form to get a copy of what we hold about you, correct it, or delete it — that is the durable route for a data request regardless of what other support channel is live at the time. We will answer within 30 days. You never need an account to use the free checker, and you never need one to use the toolkit — it exists only so you can re-download.
Cookies
The only cookie this site sets is a session cookie, and only after you sign in. There are no analytics, advertising or preference cookies.
Changes
If this policy changes in a way that affects what we collect, the date at the top changes and the previous version stays available on request.
Contact
SplitCheck — reachable through the contact form.