Skip to content

Privacy Policy

How we handle your data at Rool.

Effective date: November 9, 2026

Company: Lightpost One Ltd (trading as Rool)

Location: Dublin, Ireland

Contact: contact@rool.dev

Lightpost One builds Rool, a platform and developer toolkit built around private collaborative virtual machines ("Rool Machines") for working with structured data, files, and AI-powered tools.

We aim to be straightforward and respectful about privacy. We do not sell personal data, there are no ads in Rool, and we never use your content to train AI models. We use data only to operate, secure, support, and improve the Rool service.

This policy explains what we collect, how we use it, and your rights.

1. Who we are

Lightpost One Ltd is the operator of Rool.

We are the data controller for your account data and for the operation of the Rool service.

If you use Rool through your employer or another organisation, that organisation decides how the content in its Machines is handled, and its privacy policy applies to that content. We handle it only as the organisation instructs us. Its admins also decide which optional providers, listed in section 5, its machines use.

2. Data we collect

Account data

When you create an account, we collect:

  • Name
  • Email address
  • Sign-in information: a password (stored hashed) or passkey if you set one, and the identifier Google or Apple gives us if you sign in with them
  • Billing or subscription information (if applicable)

Service data (your Rool Machines)

This includes the content you create or upload while using Rool, such as:

  • Objects, relations, and metadata
  • Files and media
  • Prompts and interaction history with AI features
  • User storage and preferences
  • Collaboration data (members, roles, activity)

Usage and technical data

Collected automatically when the service is used:

  • IP address
  • Device/browser information
  • Timestamps
  • API requests and feature usage
  • Performance metrics and error logs

3. How we use data

We use data to:

  • Provide and operate the Rool service
  • Authenticate users
  • Store and synchronize Rool Machines
  • Enable collaboration features
  • Execute AI operations you request
  • Maintain reliability and security
  • Provide support
  • Understand how Rool is used, so we can find problems and improve the product

We do not:

  • Use your content to train AI models
  • Sell personal data
  • Show ads in Rool
  • Track your browsing on other websites, or build an advertising profile of you. If you arrive through one of our ads, we tell that ad platform about your visit and sign-up, as described in section 5, and nothing more.

Lawful basis

Here is the legal basis for each of the uses above:

  • Our contract with you (Article 6(1)(b)): providing and operating the service, authenticating you, storing and synchronising your Machines, collaboration, running the AI operations you request, and support.
  • Our legitimate interests (Article 6(1)(f)): keeping the service reliable and secure, detecting abuse, diagnosing faults, understanding how Rool is used so we can improve it, and measuring our advertising. You can object to processing on this basis, as described in section 8.
  • Legal obligations (Article 6(1)(c)): keeping billing and tax records, and responding to lawful requests from authorities.

None of this is based on consent, so there is no consent to give or withdraw. Your controls are the switches in Settings for your own account, which cover the optional providers in section 5, and the rights in section 8.

4. AI processing, support, and product improvement

When you use AI features (such as prompts, placeholders, search, or generation), relevant content and context are processed to provide the requested functionality. Section 5 lists which providers process it.

No model training

We do not use your content to train or otherwise improve AI models. This applies to your files, messages, conversations, and everything else stored in your Rool Machines. It applies to the models we run ourselves and to the AI providers in section 5, and to every plan, including the free one.

Support and quality assurance

We use usage and technical data to monitor reliability, detect abuse, and understand how features are used.

When you ask us for support, when we investigate a security incident, or when we diagnose a fault in the service, such as a failed, malformed, or incorrect AI response, we may need to access the data in your Machine. We do so only as far as that purpose requires, and where feasible we work with aggregated or de-identified data instead.

5. Data sharing

Rool is built on infrastructure and services we do not run ourselves. Each provider below does one job for us and receives only the data that job needs. The formal register, with legal entities, processing locations, and a change log, is at rool.dev/subprocessors.

Providers that run Rool

These providers are part of running Rool, so they are involved whenever you use it:

  • Hetzner (Finland) hosts Rool's servers and storage, including the servers that deliver the Rool app, receive your connection, and sign you in.
  • ScanNet (Denmark) hosts GPU servers our own AI models run on, in Denmark. It provides the hardware only. We deploy and run the models ourselves, and requests go to them, not to ScanNet.
  • Google Cloud holds the database with your account data and sign-in records, and encrypted restore points of your machines, in Google's Finland region (europe-north1).
  • Stripe handles payments. It receives your billing details when you buy a plan or credits.
  • Apple handles Sign in with Apple and purchases in the iOS app, when you use them.
  • Google (Google Account) handles sign-in with your Google Account, when you use it.
  • Resend handles email: the email Rool sends you, and the email your machine sends and receives. It stores that email, including its content.
  • Speechmatics turns your voice into text when you use the microphone. It receives the audio directly from your device and processes it on servers in the EU.
  • Matomo measures visits to the rool.dev website, without cookies.
  • Google Workspace handles our email. It holds the messages you send to contact@rool.dev, and our replies.

Optional providers

These providers each power a specific feature and are only involved when you use it. Requests are sent from Rool's servers under Rool's account, not yours. Each provider can be turned off for an account under Settings, Data and Privacy, and the setting covers every machine that account owns. Your own machines follow your settings. A machine owned by a team follows that team's settings, chosen by its admins, for everyone who works in it, including people from other accounts. Content you put in a team machine can therefore reach providers you have turned off for yourself.

  • TensorX (Ireland) is a backup for when Rool's own models are unavailable or busy, and runs some advanced or experimental models. It then receives what our own models would have: your messages, the conversation so far including the names of the people in it, account details the AI is given such as your plan and approximate location, images you attach, and the text of files the AI reads.
  • Google Cloud AI (Vertex AI) generates and edits images, on servers in the EU. It receives the description of the image you ask for, and the image you ask to edit.
  • Brave Software (Brave Search) provides web, image, and video search. It receives the search query, written by you or by the AI on your behalf.
  • SerpApi provides Google Scholar search. It receives the search query.
  • Jina AI (Jina Reader) renders web pages for the AI to read. It receives the address of the page. With Jina off, the AI reads pages directly, and some may come back incomplete.

TensorX and Google Cloud AI process data only on our instructions and may not use it to train AI models, and TensorX keeps nothing after answering. Jina does not train on requests, and may keep the page addresses it receives for a time. SerpApi keeps searches for 31 days. Brave receives search queries under its own terms rather than as our processor, and keeps them for up to 90 days. These providers see the query and Rool's account, not who you are, but a query contains whatever you or the AI put in it. That is why they are optional.

Advertising measurement

We advertise Rool on advertising platforms such as Google and Reddit. If you arrive through one of those ads, we tell that platform when you visit, when you sign up, and when you start using Rool, so we can see which ads work. We send only the click identifier the platform added to the link, and the page address. We send no name, email address, or content.

Websites and services you use through Rool

Your machine can reach the internet. When you, or the AI working for you, open a web page, search Wikipedia, or use an outside service you connect to your machine, that site or service receives what the request contains and handles it under its own terms. These are not Rool providers.

We may also share data with professional advisors or authorities where required by law.

Where a provider processes data outside the EU/EEA, the safeguards in section 6 (Data location) apply.

6. Data location

Your machines, files, conversations, and account data are stored in the European Union (Finland). Some providers may hold data outside the EU/EEA: Stripe (payment records), Apple (Sign in with Apple and iOS purchases), Google (sign-in with a Google Account), Resend (email, including its content), Google Workspace (processing of support email), Brave Software (search queries), SerpApi (search queries), Jina AI (addresses of web pages the AI reads), and advertising platforms (click identifiers).

Where a provider processes data outside the EU/EEA on our behalf, the transfer is covered by the European Commission's Standard Contractual Clauses in our agreement with that provider, or by the EU-US Data Privacy Framework where the provider is certified under it. Brave is the exception: it receives search queries under its own terms, not on our behalf, and no agreement covers them. We send each query with no identifier, so in Brave's hands it cannot be linked to you. On that basis, we treat queries sent to Brave as not being a transfer of personal data. A query may still contain whatever you or the AI typed into it, which is why Brave is one of the providers an account can turn off for its machines. You can ask us for a copy of the safeguards that apply to a particular provider at contact@rool.dev.

7. Retention

We retain data only as long as necessary:

  • Account data: while your account exists. When you delete your account, it is removed 7 days later, unless you sign in again before then.
  • Rool Machine data: until you delete it
  • Restore points: each machine keeps restore points, every one from the last hour, one per hour for the last day, one per day for the last 30 days, and one per month for as long as the machine exists. Something you delete inside a machine remains in its restore points until they expire.
  • AI records: we keep a record of each AI request and response for 30 days, to diagnose faults and investigate abuse. These records expire after 30 days even if you delete the conversation, machine, or account sooner.
  • Logs: 30 days, for operational and security purposes

Deleting a Rool Machine removes it and its restore points from active systems. Copies in our storage and database backups expire within 7 days.

8. Your rights (GDPR)

These rights come from the GDPR, and we extend them to everyone who uses Rool, wherever you are. You have the right to:

  • Access your data
  • Correct inaccuracies
  • Export your data
  • Delete your data
  • Restrict what we do with your data, or object to anything we do on the basis of our legitimate interests

Your files are yours to take at any time: download them from your Machine, or ask the AI to package them in whatever form you need.

To exercise your rights, contact: contact@rool.dev

You may also lodge a complaint with the Irish Data Protection Commission.

9. Security

We use industry-standard safeguards, including:

  • Encryption in transit (HTTPS/TLS)
  • Each machine runs in its own virtual machine, and its storage is encrypted with its own key
  • Access controls
  • Secure infrastructure
  • Monitoring and updates

No system is perfectly secure, but we work to protect your data using best practices.

10. Children

Rool is a general-purpose productivity and developer platform and is not directed to children under 16.

We do not knowingly collect personal data from children. If we become aware that a child has created an account or provided personal data, we will take steps to delete the information.

If you believe this has occurred, please contact contact@rool.dev.

11. Changes

We may update this policy as the service evolves. We will notify users of material changes.

12. Contact

Questions or requests:

contact@rool.dev
Lightpost One Ltd
Dublin, Ireland