Privacy
Short version: Kit Rx never stores your statement. To read it, the file is sent to our AI provider, Anthropic, which under its standard policy deletes API inputs within 30 days and keeps them longer only in the cases its policy lists, such as a usage policy violation or a legal requirement. Here is exactly what happens.
What happens to your PDF
- Your browser sends the PDF over HTTPS to our server, a Cloudflare Worker. Before upload, the page count is read in your browser.
- The Worker holds the file in memory for the length of that one request. It counts the pages and sends the file to Anthropic's API (the Claude AI model), which reads it and returns the transactions.
- The Worker checks the balances, sends the result to your browser, and the request ends. The file and the result are gone from our side.
We never write your file, its text, the transactions, names, addresses, or account numbers to a database, file storage, cache, or log. There is no uploads folder on our side and nothing for us to delete later. The copy Anthropic receives is covered by Anthropic's policy, below.
Anthropic (AI provider)
Statements are read by Claude through Anthropic's commercial API. The PDF travels inside the API request; Kit Rx does not use Anthropic's Files API or any other storage Anthropic offers. What Anthropic's published policies say (checked 2026-11-15):
- Retention: Anthropic automatically deletes API inputs and outputs from its back end within 30 days of receipt or generation. It can keep them longer to enforce its Usage Policy (inputs and outputs flagged by its automated trust and safety systems are kept for up to 2 years) or to comply with the law, and where a customer's contract permits, it may anonymize data for research or statistics and keep that longer.
- Training: by default, Anthropic does not use inputs or outputs from its commercial products, including the API, to train its models.
Sources: Anthropic's data retention article and training article. Anthropic's policies can change; those pages are the current word.
Your results
The transaction table lives only in your browser tab. Excel and CSV files are generated in your browser, not on our server. Closing the tab discards everything.
What we do store
- Free pages: a salted hash of your IP address and how many free pages it used that day. The salt changes every day, so old records cannot be linked back to an IP address.
- Paid access keys: a one-way hash of the key (not the key itself), the remaining credits, the plan, the Stripe customer and subscription IDs, and the email address you gave Stripe, used only to help you recover a lost key.
- Usage counts: for each conversion, the number of pages and transactions, whether it balanced, how long it took, and what it cost us. No statement content.
- Daily counters: how many times a day things like "checkout started", "purchase completed", or "free limit reached" happened. Each counter is only a date, an event name, and a number, with no IP address, key, email, or file details.
- Purchases: for each payment, the Stripe payment reference, the plan, and the pages it added, so a refund or chargeback can take those pages back.
Payments
Stripe processes payments. We receive the payment status and the email address you enter at checkout; we never see card numbers.
Hosting, cookies, and tracking
The site runs on Cloudflare, which processes standard connection data (such as IP address and browser type) to deliver and protect the site. We use no cookies, no analytics, no ads, and no third-party scripts. Your access key is kept in your browser's local storage so you don't have to type it again; remove it any time on the account page.
Contact
Questions or requests: support@kit-rx.com.
Last updated 2026-11-15.