Privacy Policy

Splitroom is an AI-powered synthetic-audience platform. Some of the data you provide is sent to third-party AI inference providers so we can generate the simulation outputs you have requested. See Section 4 — Third-party services for the current list.

1. Who we are and how to contact us

The data controller for information processed through Splitroom is Satak Technologies. If you have any questions, complaints, or requests relating to your personal information, please contact us at privacy@satak.tech.

Where applicable local law (for example the EU General Data Protection Regulation, the UK GDPR, or India’s Digital Personal Data Protection Act 2023) grants you specific rights, this Policy is intended to be read consistently with those laws. In the event of a direct conflict between this Policy and a mandatory local law, that mandatory local law prevails.

2. Information we collect

Information you give us

Account information: email address, display name, authentication tokens issued through Google Firebase.
Creative material: images or other creative assets you upload for simulation.
Audience definitions: the demographic, psychographic, and category parameters you configure for a simulation.
Communications: emails, support tickets, and other messages you send us.

Information we collect automatically

Technical information: IP address, device identifiers, browser type, operating system, timestamps, referrer URLs.
Usage information: pages viewed, actions taken in the dashboard, features used, error reports.
Cookies and similar technologies: see our Cookie Policy for a full description.

Simulation outputs

When you run a simulation, we store the resulting report, including alignment scores, dimension distributions, synthetic panelist reactions, debate transcript, and metadata (such as which model was used, the engine version, and how long the run took). We treat these outputs as your content for the purposes of storage and access controls.

Account memory

Your workspace's own history is used to make your own future runs sharper: prior verdicts for a brand may be provided as context to the analysis of that brand's later simulations, saved synthetic panels may be reused when you test the same audience brief again, and outcomes you voluntarily report back (which creative actually won in-market) are stored alongside the relevant simulation. Account memory is private to your workspace. It is not shared across customers, and it is not used to train any model.

Model training

We do not use your uploaded creatives or your simulation outputs to train machine-learning models. If we ever propose to use customer data for model improvement, it will be on an opt-in basis, limited to aggregated and anonymized data, described clearly at the point of consent, and announced to you before it takes effect. Until you have opted in, nothing changes for your account.

3. How we use information

We use the information above for the following purposes:

To provide, operate, and improve the Service.
To authenticate you and secure your account against unauthorised access.
To generate simulation outputs you have requested.
To monitor Service health, diagnose issues, and prevent abuse.
To communicate with you about the Service, security incidents, changes to these documents, and (with your consent, where required) product news.
To comply with legal, tax, and regulatory obligations.
To defend or enforce our legal rights, including these Terms and our Acceptable Use Policy.

Legal bases (where applicable)

Where the EU or UK GDPR applies, we rely on the following legal bases: performance of a contract (Article 6(1)(b)) to operate the Service; legitimate interests (Article 6(1)(f)) to secure the platform, prevent abuse, and improve the product; legal obligation (Article 6(1)(c)) where processing is required by law; and consent (Article 6(1)(a)) where we ask for it (for example, for optional marketing communications).

4. Third-party services and international transfers

To operate the Service, we share information with the following categories of third-party processors:

Cloud infrastructure & storage: Google Firebase (Alphabet Inc. and its affiliates) for authentication, database, storage, and hosting.
AI inference: OpenRouter, Inc. and, through it, model providers including Anthropic, PBC and OpenAI, OPCO LLC (and their affiliates). When you run a simulation, prompts derived from your creative material and audience configuration are sent to these providers for processing.
Payment processing (when applicable): we will name our payment processor on the pricing page before you transact.
Analytics and error reporting: product analytics providers used to understand aggregate usage and diagnose issues, kept to the minimum necessary.

These providers may be located outside of your country of residence, including in the United States and other jurisdictions. Where required, we rely on Standard Contractual Clauses or other appropriate safeguards for international transfers.

We do not sell personal information. We do not share your account information with advertising or marketing networks for cross-context behavioural advertising.

5. How AI inference providers handle your data

The AI inference providers we route requests through have their own privacy commitments and terms. As of the last update to this Policy, the providers we currently rely on for inference do not train their foundation models on inputs sent through their commercial APIs by default. Nonetheless, once a prompt leaves our systems, it is subject to the provider’s own privacy and security controls.

If you do not want your creative material or audience definitions processed by third-party AI providers, do not use the Service, as providing simulation outputs is not possible without such processing.

6. How long we retain information

We retain account information and simulation outputs for as long as your account is active, plus a limited additional period for backup, legal, and audit purposes.

You may delete individual simulations and the creative content associated with them from within the dashboard. When you delete a simulation, we remove the primary record; backup copies persist for up to ninety (90) days before being purged from our backup rotation.

When you close your account, we delete or anonymise your personal information within a reasonable period, subject to any legal obligation to retain records (for example, tax or accounting).

7. Security

We use commercially reasonable administrative, technical, and organisational measures to protect information against loss, misuse, and unauthorised access. These measures include encryption in transit (HTTPS), access controls, principle-of-least-privilege for internal systems, and infrastructure operated by reputable providers.

No method of transmission or storage is 100% secure. If you suspect your account has been compromised, contact us immediately at security@satak.tech.

8. Your rights

Depending on your jurisdiction, you may have the following rights in relation to your personal information:

The right to access the personal information we hold about you.
The right to correct inaccurate or incomplete information.
The right to delete your personal information, subject to legal or legitimate-interest exceptions.
The right to export a copy of your personal information in a portable format.
The right to object to or restrict certain processing, or to withdraw consent where processing is based on consent.
The right to lodge a complaint with your local data-protection authority.

To exercise these rights, contact us at privacy@satak.tech. We will respond within the time limits set by applicable law (typically 30 days).