Riplo application privacy notice

Last updated 5 October 2026

This notice covers the Riplo application at os.riplo.ai: the decks you create, the personal data in your account, how AI features process your content, and what happens when you connect an external AI assistant. It supplements the Riplo Limited privacy policy (opens in a new tab), which covers our website and prospective customers. Where the two differ about the application, this notice governs.

Riplo Limited is the data controller, registered in England and Wales, company number 16902933.

1. What we process

Account and organisation data, from your sign-in provider:

  • Name, email address, profile image, and the organisations and teams you belong to, together with your role in them.
  • Authentication is handled by WorkOS AuthKit. We never receive or store your password.

Content you create in the product:

  • Decks, slides, text, tables, charts, speaker notes, comments and annotations, themes and templates, and the version history of those decks.
  • Images and other files you upload, plus files we generate from your decks such as PowerPoint exports and rendered slide images.
  • Any personal data you choose to put inside a deck. Slides are free-form, so please treat them as you would any other business document and avoid placing special-category or unnecessary personal data in them.

AI interaction data:

  • The messages you send to Riplo's AI features, the model's replies, and a record of the actions it took on your deck.
  • Requests made by external AI assistants you connect, described in “Connected AI assistants”.

Operational data:

  • Product analytics events (which features are used, and by which account), error diagnostics, and audit records of significant actions such as sharing or deleting a deck.
  • Email records: which of our emails you have chosen to receive, a record of each change to that choice (when, and where you made it), and delivery records for the emails we send you, such as whether one was delivered or bounced.

2. Why we process it, and our lawful basis

  • To provide the service — storing and rendering your decks, running AI features you invoke, and enforcing who can open what. Basis: performance of our contract with you or your organisation.
  • To email you about the service — sign-in codes, invitations, a welcome email, and notices when someone shares a deck with you or adds you to a team. Basis: performance of our contract with you or your organisation.
  • To send you product updates and tips — occasional emails about new features and ways to use Riplo. Basis: legitimate interests, relying on the “soft opt-in” for people who hold a Riplo account. When you sign up we offer a box to opt out before you continue; if you held an account before we began sending these emails, this notice and every email tell you how to stop them. See Emails we send.
  • To keep the service secure and reliable — error diagnostics, audit records, abuse prevention. Basis: legitimate interests.
  • To improve the product — understanding which features are used and where they fail. Basis: legitimate interests.
  • To meet legal obligations — responding to lawful requests and keeping records we are required to keep. Basis: legal obligation.

We do not sell your personal data, and we do not use your deck content or prompts to train our own models or anyone else's.

3. How AI features process your content

When you use an AI feature, the relevant part of your deck and your instruction are sent to a large language model so it can answer or make the change you asked for.

  • Models run on Amazon Bedrock in the eu-west-2 (London) region, through EU inference profiles. Your content is processed inside AWS under our account.
  • AWS does not use content submitted to Bedrock to train models, and Bedrock does not retain it after the request.
  • Some research features can retrieve information from the public web through a search provider. Only your search query is sent, not your deck.

We keep our own record of AI activity so you can see what changed and so we can debug failures. See Retention for how long each part is kept.

4. Connected AI assistants

You can connect an external AI assistant — such as Claude or ChatGPT — to Riplo over the Model Context Protocol (MCP). This is worth understanding before you enable it:

  • You connect as yourself. The assistant signs in with your Riplo account and can only reach the decks you can already open. It gains no additional access.
  • Content flows to the assistant's provider. When the assistant reads a deck, that content is sent to it, and is then handled under your agreement with Anthropic, OpenAI, or whichever provider you chose — not under this notice. Please check their terms and retention practices.
  • We record the call, not the deck. For each request we store which tool ran, which operations it performed, timings, and the outcome. When a request fails or is rejected we also store the request and response so the failure can be diagnosed; those records are size-capped. Successful requests store no deck content.
  • You can disconnect at any time from the assistant, which immediately revokes its access.

5. Emails we send, and how to unsubscribe

Sign-in codes, password resets and invitations are always sent, because the service cannot work without them. Everything else can be turned off:

  • Sharing and teams: when someone shares a deck with you or adds you to a team.
  • Getting started: your welcome email and a few pointers while you are new.
  • Product updates, and tips and ideas: occasional news about Riplo. Guests invited to someone else's workspace never receive these.

To stop any of them:

  • use the unsubscribe or “manage email preferences” link at the foot of every email, which works without signing in;
  • change your choices in the application under Settings → Notifications; or
  • write to support@riplo.ai.

We keep a record of each choice you make, so we can show that we only email you as you asked.

6. Who we share it with

We use a small number of service providers, each under a data processing agreement and each limited to what its function requires:

  • Amazon Web Services — database, file storage, cache, and AI model inference. London (eu-west-2).
  • Vercel — application hosting. Served from London.
  • WorkOS — authentication and organisation directory.
  • Resend — email delivery. Emails are sent from Ireland (eu-west-1); Resend stores account data, including your email address, name and email preferences, in the United States.
  • PostHog — product analytics. EU hosting.
  • Sentry — error and performance diagnostics. EU hosting.
  • Grafana Labs — log and trace storage for operational monitoring. UK hosting.
  • A web-search provider and a company-logo lookup provider, used only for the specific features that call them.

We also disclose data where the law requires it, and to professional advisers where necessary. For a current list of sub-processors, write to support@riplo.ai.

7. Diagnostics and session recording

To find and fix faults, our error-monitoring provider captures diagnostic information when something goes wrong, including the account involved and the state of the page at the time. A sampled subset of sessions, and sessions in which an error occurs, include a replay of the interface.

Because Riplo is a document editor, a replay can show the deck that was open. Replays are used only to diagnose faults, are visible only to Riplo staff who need them, and are retained by the provider on a limited window rather than indefinitely. If you would prefer your organisation excluded from replay capture, contact us at support@riplo.ai.

8. Where your data is stored

The application, its database, file storage, cache, and AI inference all run in the United Kingdom and the EEA — Vercel in London, and AWS in London (eu-west-2). Databases and file storage are encrypted at rest, and traffic to the application is encrypted in transit.

Some providers listed above may process limited operational data (such as diagnostics) outside the UK and EEA, and our email provider stores your email address, name and email preferences in the United States. Where that happens, we rely on UK International Data Transfer Agreements, the EU standard contractual clauses, or an adequacy decision.

9. Retention

We keep different things for different periods, and we would rather be precise than reassuring:

  • Decks and uploads — kept until you or your organisation delete them. Deleting a deck hides it immediately and removes it from the product; the underlying record may persist until the project or organisation is deleted.
  • Version history — automatic snapshots are thinned over time and removed after 30 days. Snapshots you create deliberately are kept.
  • Chat transcripts and the assistant's working memory — kept for as long as the conversation exists, so you can return to it. Deleted when the project or organisation is deleted.
  • Detailed AI run records (the full record of what a model was sent and returned, used for debugging) — deleted automatically after 30 days.
  • Exports and rendered images — deleted automatically after 30 days.
  • Emails we send — the details an email was built from are cleared after 90 days, and the delivery record is deleted after 13 months. The record of your email choices is kept as proof of what you agreed to; you can ask us to erase it.
  • Connector request records and audit logs — kept while your organisation is active, because they are what lets an administrator review after the fact who changed what.
  • Account records — when someone leaves an organisation we retain a minimal record of their identity so that their past work, share grants, and audit entries remain attributable. You can ask us to erase it; see Your rights.

10. Cookies and local storage

  • Session cookie — an encrypted cookie that keeps you signed in. Strictly necessary; the product cannot work without it.
  • Analytics cookies — set by our analytics provider to recognise a returning browser and group events into sessions.
  • Local storage — small interface preferences such as which panel you had open and the slide you were last on. This stays in your browser.

11. Your rights

Under the UK GDPR and equivalent EEA law you may request access to your personal data, correction of inaccurate data, erasure, restriction of processing, portability, and you may object to processing based on legitimate interests.

If you use Riplo through your employer, that organisation controls the workspace and its content; we will normally direct requests about workspace content to them and assist as their processor.

To exercise a right, email support@riplo.ai with “data request” in the subject, or write to our Data Protection Officer at the postal address below. Requests reach the same place either way. We respond within one month. You may also complain to the UK Information Commissioner's Office, or your local supervisory authority in the EEA.

12. Children

Riplo is a business product, is not directed at children, and is not available to anyone under 18. We do not knowingly collect data from children.

13. Changes and how to contact us

If we change how we handle your data we will update this notice and its “last updated” date, and will tell you in the product where the change is significant.

Data protection enquiries and product support both reach us at support@riplo.ai, which is also the address for our Data Protection Officer, Zack Zornitta. By post: Riplo Limited, 167-169 Great Portland Street, London, W1W 5PF, United Kingdom.