Legal

Privacy Policy

Last updated: September 2026

This policy explains what RubusArc (“we”, “us”) collects, why, and your choices. RubusArc at mabel.chat is four things on one platform: a website assistant, AI-readiness assessments (Assess), structured collection (Collect), and document processing (Back office).

A workspace is normally one of those. A Portfolio workspace is two of them at once: an investor — a fund, a family office, a corporate development team — assessing and collecting from the companies it holds, in one place rather than one workspace per company. Nothing in this policy is different for it. Everything below about Assess and about Collect applies to a portfolio workspace unchanged.

Who is responsible for what. For your account we are the data controller. For everything the service handles on your behalf — visitor conversations, stakeholder interviews, collected responses and documents, the contents of a mailbox you connect — you are the controller and we are your processor.

Assess and Collect often add a third link. When a consultant uses RubusArc to interview their client’s staff, the client is the controller, the consultant is their processor, and RubusArc is a sub-processor. We have no relationship with the client and no way to reach them: if you were interviewed or asked to submit information and want to know why, ask the organisation that sent you the link — we will help them answer, but the answer is theirs.

An investor assessing a company in its portfolio is the same shape and gets the same answer. The people interviewed work for the company; the reason they were asked is a relationship between that company and its investor, not one either of them has with us. We hold the interviews and the responses on the investor’s instructions, and we cannot tell you why you were asked — the organisation that sent you the link can.

What we collect

• Account data — your email and name, workspace and business details, and sign-in records.
• Website content — the public pages of the site you connect, used to build and ground your assistant.
• Conversations — chat transcripts and voice audio/transcripts between your assistant and your visitors, plus any leads it captures (such as name, email, or phone).
• Assessment data — the people you nominate as stakeholders (name, work email, role, team), the interviews they hold with the AI interviewer, the findings and verbatim quotes recorded from those interviews, and any documents you upload to describe the engagement. Reports quote stakeholders by name and role, because a claim nobody can attribute is a claim nobody can check.
• Collected responses — whatever your campaign asks for, which is your choice and can include identity numbers, passport or registration documents, and financial details. Fields you mark sensitive are encrypted at rest with a key we hold separately, and are never shown in exports or totals.
• Mailbox and drive contents — for Back office, the messages and attachments matching the filters you set on an account you connect, and the documents extracted from them. Filters narrow what is processed; the connection itself reaches the account.
• Usage & technical data — logs, device/browser info, and IP addresses, used for security, reliability, and abuse prevention.
• Billing data — subscription status and payment records handled by our payment processor; we don’t store full card numbers.

How we use it

To provide and improve the service: build, host, and run your assistant; answer and log conversations; conduct interviews and compile the reports that cite them; collect and reconcile the responses your campaigns ask for; read the documents you point us at and file them into your ledger; process payments; provide support; and keep the service secure. We don’t sell your data or use it for third-party advertising.

We do not train AI models on your data, and our model providers are engaged under terms that do not permit them to either.

Automated decisions

RubusArc writes, groups and prices; it does not decide. A report’s recommendations are for a human to accept or decline, an extracted document is held for a person to review before it counts, and a compliance decision is recorded against the named person who made it. No outcome that affects an individual is produced by the service without someone acting on it.

Cookies

We use a small number of essential cookies — mainly to keep you signed in. We don’t use advertising or cross-site tracking cookies.

Service providers

We share data only with providers that help us run RubusArc, under contracts that protect it — for example: cloud hosting; AI model providers (for generating answers, conducting interviews, and reading documents); voice transcription/synthesis; website crawling; email delivery; sign-in via Google; document and mailbox access with the accounts you connect; company-registry lookups where a campaign enables them; and payment processing. They may only use the data to provide their service to us, and may not train on it.

We may also disclose data if required by law, to protect rights and safety, or as part of a business transfer (with this policy continuing to apply).

Retention

We keep account data while your workspace is active. Visitor conversation data is kept according to your workspace’s retention setting and then deleted. You can delete conversations, leads, or your whole workspace at any time, and visitors can ask your assistant to delete their data.

Two exceptions matter, and both are deliberate:

• Interview transcripts are not swept on that clock. They are the evidence behind a report’s citations — deleting them on a timer would turn a cited report into an uncheckable one. They are removed when the engagement or workspace is deleted.
• A recorded retention obligation is a floor, not a target. Where a collection campaign stamps a retention period on a reviewed record — a compliance check that must be kept for a set number of years — that record survives an ordinary delete until the period ends. Nothing recomputes that period afterwards, so a later configuration change cannot silently shorten an obligation already recorded.

What we do not offer

Responses are attributed. Answers given through a collection campaign, including staff surveys, are recorded against the person who gave them and are visible to the organisation running the campaign. RubusArc does not currently provide anonymous collection, and no survey run on it should be presented as anonymous.

Your rights

Depending on where you live, you may have rights to access, correct, export, or delete your personal data, and to object to or restrict certain processing. Contact us to exercise them.

Because most of what we hold is processed on behalf of a business, the right place to ask is usually that business: the operator of the assistant you spoke to, or the organisation that sent you an interview or collection link. We will help them respond. If you don’t know who that is, write to us and we will route it.

Security

We protect data with encryption in transit, strict tenant isolation enforced in the database itself rather than only in application code, access controls, and regular backups. Sensitive collected fields and the credentials for any mailbox or drive you connect are additionally encrypted at rest and are never returned in readable form. Administrative access requires two-factor authentication. No system is perfectly secure, but we work to keep your data safe and will notify you of significant incidents as required by law.

International transfers & children

Your data may be processed in countries other than your own, with appropriate safeguards in place. RubusArc is not intended for children, and we don’t knowingly collect data from anyone under 16.

Changes & contact

We may update this policy and will note the date above; material changes will be communicated. Questions or requests? Email privacy@mabel.chat.