Privacy policy for Serendipity
Serendipity by Valleyport
Privacy Policy
Last updated 28 October 2026. Serendipity is operated by Valleyport Ltd (company number 17361820), registered in England & Wales at 181 Mitcham Lane, London, SW16 6NA.
What we collect
Serendipity does not use accounts, logins, names, or email addresses. On first launch, your device is given a random pseudonymous ID, stored only on your device and used to identify your data on our server. We collect: the pages you like, dislike, save, or underline; a taste profile built from that activity so the app can personalise what it shows you; and basic session logs (what was served, when) used to keep the feed varied and to run the app's own internal usage metrics. We also process topic-search text to find and prepare relevant pages. Topic searches may be cached to improve later searches, but are not stored in that cache alongside your pseudonymous user ID.
When you follow one of our campaign download links, we record coarse campaign metadata such as source, campaign, and creative to measure whether the link led to an install. We do not retain the raw browser Referer URL or the raw Google Play Install Referrer string.
If you deliberately suggest a webpage for the shared corpus, we retain its normalised URL with your pseudonymous profile for up to 30 days to prevent duplicate or abusive submissions. Do not suggest private, login-only, or token-bearing links. A suggestion is reviewed through the normal safety and quality pipeline; it does not automatically become visible to anyone.
We do not collect your name, contacts, precise location, or general browsing history outside the app, and we do not embed any third-party advertising or analytics trackers in the app.
External pages and website data
Serendipity displays one selected external webpage at a time. The selected website and resources it embeds receive ordinary web-request information from your device, including your IP address and browser User-Agent. External sites operate under their own privacy practices and may use their own first-party cookies, browser storage, trackers, or third-party resources. Serendipity does not control those independent technologies. To make Surf responsive while the app is open, Serendipity may preload up to three of the next queued pages in the background on a suitable Wi-Fi device (and fewer on lower-memory or metered connections). Those pages and their embedded resources receive the same ordinary web-request information and may set their own cookies or browser storage before you choose to view them.
How we use it
Your activity is used only to run the feed itself: personalising what's served next, syncing your data across your own devices if you use the pairing-code sync feature, and improving the quality of the shared content corpus for everyone. It is never sold, and never shared with advertisers.
Third parties involved in running the service
Some processing happens through third-party infrastructure providers acting on our instructions, not as independent users of your data: Cloudflare (secure traffic delivery and abuse protection), Railway (application, database, and queue hosting), GitHub (private database-backup artifacts), OpenAI (generates embeddings and screens topic queries and page text for safety), Anthropic (expands topic searches and scores borderline content for quality), and Brave Search (finds candidate pages for live topic searches with strict SafeSearch enabled), Sentry (receives narrow crash and error diagnostics so we can find and fix bugs, including exception stack, app/build version, and device/operating-system context; request URLs, topic text, page IDs, request bodies, and your Serendipity pseudonymous ID are excluded), and Resend (delivers an email notification containing a random feedback-row reference, but not the feedback message itself). None of these providers receive your name, email address, contacts, or precise location, these providers receive the text of optional product feedback except Railway, where the profile-linked feedback row is stored. They receive only the information needed for the relevant operation; Brave Search also receives standard network request information from our server.
Content moderation and safety
New pages entering Serendipity pass through an automated safety and quality pipeline before they become available: a static block-list for known adult/gambling/piracy sources, automated text moderation, and quality scoring. Live topic search additionally uses strict SafeSearch and blocks prohibited topic-search intent. Older corpus pages are being checked through the same resumable safety process. The in-app report button immediately removes a reported page from that reporting profile and sends it to a manual review queue. A trusted moderation decision is required for a global quarantine. These measures reduce but cannot eliminate the possibility of unsuitable content because Serendipity links to the changing open web rather than hosting the underlying pages.
Browser extension
The Serendipity browser extension (for Chrome, Microsoft Edge, Firefox, and other Chromium-based browsers such as Brave and Opera) works the same way as the app: a random pseudonymous device identifier is generated and stored only in the extension's local browser storage, never in a cookie or synced through your browser account. The extension's background script reads the address of a page only in a browser tab where you have started Serendipity, so it can keep its side panel, sidebar, or window attached to that tab as you browse; it does not read, store, or transmit the addresses of your other tabs, and it never reads a page's contents, only its address once it finishes loading. That locally observed address is sent to our server only when you explicitly press Save on a page you navigated to, so it can be added to your Finds and, if it isn't already part of the curated corpus, considered for it through the same safety and quality pipeline described above. Liking, disliking, exploring, reporting, and viewing a recommendation do not transmit any browsed address. Topic searches you type in the extension are sent to our server and handled like topic searches in the app. In Firefox, the extension asks for your permission to connect to our server the first time you open it. The extension requests the narrowest browser permissions its features need and does not request permission to read or modify the content of the pages you visit. The same 30-day raw-event retention, in-app-equivalent data export, and permanent-deletion guarantees described elsewhere in this policy apply to data collected through the extension, and a pairing code lets you link an extension installation to an existing app profile without creating a second one. Serendipity's use and transfer of information received through the Chrome Web Store adheres to the Chrome Web Store User Data Policy, including its Limited Use requirements.
Advertising and sponsored content
The current version of Serendipity contains no advertising or third-party advertising SDK. If clearly-labelled sponsored pages are introduced in a future version, this policy and the relevant app-store disclosures will be updated before that feature is enabled.
Data retention and deletion
Raw records of pages served are automatically deleted after 30 days. Likes, saves, feedback, highlights, category controls, profile vectors, session summaries, and optional product feedback are kept while the pseudonymous profile remains so its feed and personal features continue to work. Globally cached topic labels are not linked to a profile and are refreshed or removed when no longer operationally useful. URL-submission records are deleted after 30 days.
You can permanently delete your profile-linked data at any time from within the app (About & privacy → Delete my data). This immediately and permanently removes your taste profile, likes, saves, highlights, reports, category feedback, product feedback, and session history from the live database, and your device generates a fresh pseudonymous identity afterward. Nightly private database backup artifacts expire after 30 days, so an erased row can persist in a historical backup until that artifact expires. Security/hosting logs and narrowly filtered Sentry crash diagnostics follow the operational retention settings of the relevant processor and are not used to reconstruct a deleted profile. Email notifications contain only a feedback-row reference, not the submitted message.
Deleting your Serendipity profile does not reliably clear cookies, cache, or browser storage previously created by independent external websites inside the Android WebView. You can clear the app's storage through Android settings (or reinstall it) to remove that local website state.
Children's privacy
Serendipity is not directed at, and is not intended for use by, anyone under the age of 16. We do not knowingly collect data from children under 16. Because we don't collect names, emails, or other identifying information, we have no way to verify a user's age directly — see the app's age-assurance documentation for how this is being addressed as the service grows.
Your rights (UK GDPR)
Because Serendipity is pseudonymous, most requests are best made through the in-app delete function, which is immediate and requires no correspondence. If you have a request our in-app tools can't handle, or a complaint, contact us at support@valleyport.co.uk. You also have the right to lodge a complaint with the UK Information Commissioner's Office (ico.org.uk).
International transfers
Our infrastructure and processors (Cloudflare, Railway, GitHub, OpenAI, Anthropic, Brave Search, Sentry, and Resend) may process data outside the UK, including in the United States. Where this happens, it's covered by those providers' standard data protection safeguards.
Changes to this policy
We'll update the date at the top of this page when this policy changes. Continued use of the app after a change means you accept the updated policy.
Contact
Valleyport Ltd, 181 Mitcham Lane, London, England, SW16 6NA. support@valleyport.co.uk