Privacy Policy
Effective date: 16 October 2026. Published: 8 October 2026.
This notice is published in advance. Until the effective date, see the previous platform privacy policy and previous SquadraOne privacy policy. An older website privacy policy is also preserved for reference.
1. Who is responsible
The service operator is Eric Leibenguth, entrepreneur individuel (EI), trading as Squadra. Contact for privacy matters: contact@squadra.ai. Address and registration details are provided in the Legal Notice.
This notice covers the Squadra website, agent-building platform, APIs and SquadraOne.
The operator is controller for account administration, billing, service and promotional communications, website analytics and its own security processing. For personal data processed within a customer’s organization on that customer’s instructions, the customer generally acts as controller and Squadra as processor under the DPA. If the customer is itself a processor, Squadra may act as a further processor. Actual purposes and instructions determine the role.
If you interact with an agent deployed by another organization, that organization’s privacy notice also applies. It determines the agent’s purposes, configuration, connected tools and retention settings. Please contact it about its handling of your data; we assist it with requests where required.
2. Data, sources and purposes
We receive data directly from account holders and Users, from organization administrators, from authorized connected tools, and from technical operation of the service. Depending on agent configuration, Customer Data may also include information about other people obtained from documents, knowledge bases or public sources selected by the customer.
| Data category | Examples | Purpose and basis where we act as controller |
|---|---|---|
| Account and organization information | Name, email, authentication identifiers, memberships and roles | Account and service administration: contract for individual customers; legitimate interests in administering business customer accounts for their representatives and Users |
| Subscription and billing information | Plan, subscription status, billing details, Stripe payment references | Contract performance and legal accounting obligations |
| Support and communications | Emails, support requests, information supplied to diagnose an issue | Contract performance or legitimate interests in supporting Users and maintaining the service |
| Promotional communications | Name, email address and communication preferences | Information about Squadra products and offers, subject to your preferences: consent where required, or legitimate interests for professional communications relevant to your activity where permitted |
| Technical and security information | IP address, browser/device information, request timestamps, error and security logs | Legitimate interests in reliability, fraud prevention and protection of the service and Users |
| Website and app analytics | Online identifiers, page/feature usage and events collected through Google Analytics | Consent where required for non-essential tracking; analysis of usage to understand service performance |
Customer Data processed on an organization’s behalf can include agent configuration, prompts, organization knowledge bases, conversations, uploaded files, action logs, personal information in connected tools, and connection credentials. We process it to provide the configured agents, execute authorized actions, store and retrieve content, troubleshoot and secure the service, and follow the customer’s instructions. The organization determines the legal basis for that processing.
You can opt out of promotional emails at any time, free of charge, using the unsubscribe mechanism in each promotional email or by contacting us. Opting out does not stop necessary service, billing or security communications.
Information necessary for authentication, billing or an authorized integration is required for the corresponding functionality; without it, that functionality cannot operate. Optional support details and non-essential analytics are not required for basic access. Do not submit unnecessary personal data or sensitive information to agents. The service is intended for professional use and is not directed to children.
3. Who can access organization data
Organization owners and administrators can access stored organization data, including conversations and files, for support and administration. Knowledge bases are shared at organization level and are not private to an individual end user. Users should avoid placing information in a shared knowledge base that other authorized organization users should not access.
API keys are stored as secrets, are not displayed after being set, and are inserted into requests by the server. End-user connection tokens are not available to organization owners or administrators. Holding an administrative role does not itself allow them to use tools connected by an end user. Agents may use those connections according to granted permissions and the configured workflow.
The service operator has administrative access to the infrastructure and may access customer data where necessary for support, maintenance, troubleshooting, security or legal compliance. This technical access is separate from organization roles. Content from customer-managed platform organizations is not used for our own product improvement or model training. SquadraOne retains the specific improvement permission explained in section 8. Analytics about service usage are distinct from customer conversation and file content.
4. AI providers and connected tools
To generate responses and execute agent workflows, relevant prompts, conversation history, instructions, knowledge-base excerpts and file content may be transmitted to AI providers. Depending on the input, this can include personal data. Do not assume automatic removal of personal information before transmission. The principal inference providers are OpenAI and Google Gemini.
-
OpenAI API: API content is not used to
train or improve OpenAI models by default, unless the API
account holder opts in. Squadra uses the Responses API
with
store: false, rather than persistent response storage. This does not eliminate abuse-monitoring logs, which ordinarily may contain content for up to 30 days, subject to legal and safety exceptions, or applicable temporary caching and feature-specific processing. See OpenAI data controls. - Gemini API: Google’s paid-service data terms exclude use of prompts and responses to improve its products; these data-use terms also apply to users in the EEA, Switzerland and UK when using unpaid services. Its Gemini API abuse-monitoring policy specifies retention of prompts, context and outputs for 55 days, separate from caching, with authorized review of flagged content. Abuse-monitoring data may be used for models specifically dedicated to policy enforcement. See Gemini terms and abuse monitoring.
Provider processing is subject to the API product, account settings and enabled features. Customers supplying their own provider credentials must check their account’s terms and settings. Additional features such as search grounding can involve additional recipients or retention; they must be considered when configuring an agent. We do not promise that inference providers never store data.
Connected tools receive the information needed for authorized actions. They may act as independent controllers or as providers to the customer, depending on the relationship. Revoking a connection prevents subsequent authorized access through that connection but does not erase data already held in the external service or undo actions already completed.
5. Recipients and international transfers
The service provider list identifies infrastructure, AI, email, analytics and payment providers, and additional providers associated with SquadraOne. Core platform infrastructure uses Google Cloud Platform, including Cloud Run, Cloud SQL, virtual machines and Cloud Storage. Database backups are stored in the EU multiregion. The website uses Firebase Hosting. Brevo delivers transactional emails, Google Analytics provides usage analytics, and Stripe processes payments. We also disclose information to professional advisers subject to confidentiality and authorities where legally required. We do not sell personal data.
EU hosting of the database does not mean all processing remains in the EU. AI inference, support and other provider operations can involve transfers outside the EEA. Where required, transfers must be covered by an applicable adequacy decision, Standard Contractual Clauses and any necessary supplementary measures, or another lawful mechanism. Contact us for information about the safeguards applicable to a particular transfer and copies where available, subject to appropriate confidentiality protections.
6. Retention and deletion
| Data | Retention period or criterion |
|---|---|
| Conversations and uploaded files | Organization-owner configuration; 12 months by default |
| Organization knowledge bases, agent configuration and connection credentials | While required for the organization’s service or connection, subject to deletion instructions and organization closure |
| Account and membership records | While the account or membership is needed, followed by only the records necessary for legal obligations, disputes and security |
| Billing records | For the periods required by applicable accounting and tax law |
| Support and security records | For as long as necessary to resolve the request, maintain security or establish, exercise or defend legal claims; relevant records may be preserved for a specific incident or dispute |
| Customer Data when an organization is deleted | Deleted from active service storage as part of organization deletion; recovery copies expire separately as described below. Customers should retrieve needed data before deleting the organization. |
| Platform data after subscription cancellation | Cancellation and organization deletion are separate operations. We do not guarantee continuing access or storage after the subscription ends; return/deletion instructions are handled under the DPA where applicable. |
| SquadraOne data after an individual user's service terminates | May be retained for up to 90 days, as in the previous SquadraOne terms, subject to earlier deletion where feasible and lawful, shorter configured retention, or legally required longer retention. This does not delay organization deletion. |
| AI-provider copies | Provider-specific retention described above; separate from the organization’s retention setting |
The SQL database retains seven automated backups and seven days of point-in-time recovery logs. Deleted records can remain in recovery copies until those copies expire through backup rotation. This is not a guarantee of permanent erasure exactly seven days after deletion. Any separately retained recovery copy must be restricted to its recovery or legal purpose, rather than used for ordinary access, and deletion instructions must be reapplied if data is restored.
Cloud Storage soft delete retains deleted files for seven additional days before permanent deletion through that mechanism. Object versioning, bucket retention policies and lifecycle rules are not enabled on the reported file-storage configuration. Application retention controls trigger deletion of active files.
Deleting an individual user account does not necessarily delete data held by an organization about that user. Data in an external connected tool is governed by that tool and the relevant organization or account holder. Contact the appropriate organization for instructions, or contact us if you need help identifying it.
7. Cookies and analytics
We use essential storage for service operation and consent choices, and Google Analytics on the website and relevant app features. Where consent is required, non-essential tracking must remain disabled until consent is given. You can refuse or withdraw consent using the available cookie controls without affecting the lawfulness of prior consent-based processing. See the Cookie Policy. Withdrawal of analytics consent does not disable essential account or security processing.
8. Specific provisions for SquadraOne
SquadraOne operates inside a dedicated Squadra organization managed by the service operator. Invitations are closed and it is maintained for existing users. The operator configures its agents and administers the organization; authorized support access therefore includes the organization administration described above as well as technical infrastructure access.
SquadraOne agents can access connected tools, such as an end user's Notion workspace, within the granted permissions. They may read, create or modify content. As administrators of SquadraOne, we configure important write or modification actions to require user approval before execution; reading actions do not require a separate approval. Platform customers independently configure approval requirements for their own agents.
For account, subscription, organization administration and our own support and security purposes, the operator acts as controller. Where an agent processes a professional user's business content on that user's instructions, the user determines those business purposes and we act as processor under the DPA. Managing SquadraOne's technical configuration does not by itself make us controller of all business content. SquadraOne uses the same storage and recovery mechanisms; its organization settings determine the actual conversation and file retention period.
The previous SquadraOne terms' permission to use User Content and Outputs to improve SquadraOne functionality, prompts, workflows and user experience is preserved. It does not extend to another customer's platform-organization content or authorize AI-model training. This is permission for a described purpose, not a statement that content has already been reused for it. Any personal-data processing for that purpose must respect confidentiality, the applicable legal basis and controller/processor roles. If we process personal data for our own improvement purpose rather than solely on a customer's instructions, we act as controller for that processing; any required authorization, compatibility assessment and information must be in place before reuse. Organization ownership alone does not remove those requirements. No unrestricted secondary use is authorized by this notice.
9. Your rights and complaints
Subject to the applicable conditions, you can request access, correction, erasure, restriction and portability of your personal data, object to processing based on legitimate interests, and withdraw consent where processing relies on it. Send requests to contact@squadra.ai. We may request proportionate information to verify identity. We normally respond within one month; a lawful extension will be explained within that period.
If we process the data for another organization, we will direct the request to it or assist it according to its instructions. You may lodge a complaint with the CNIL, the French supervisory authority, at cnil.fr, or with another competent supervisory authority. You do not have to contact us before exercising that right.
Agents generate outputs and may execute actions according to customer configuration. Customers deploying automated decisions with legal or similarly significant effects are responsible for the additional legal basis, information and safeguards required for that use. This notice does not authorize unrestricted automated decision-making by customers.
10. Security and changes
Security measures include organization roles, restricted administrative access, server-side handling of connection secrets, managed infrastructure protections and database recovery mechanisms. No internet service can guarantee absolute security. The DPA describes processor obligations, including incident notification and assistance.
Material changes to this notice will be communicated to affected account holders, normally before they take effect. We will obtain consent where required rather than treating continued use as consent to new processing purposes. Historical versions remain accessible below.
