Data Processing Agreement

For personal data contained in Customer Content (data you store or process in sandboxes, apps, functions, and managed databases), the customer is the controller (or a processor acting for another controller) and PandaStack is the processor. For account, billing, and usage data described in our Privacy Policy, PandaStack acts as an independent controller.

Subject matter and nature. Provision of isolated compute (microVM sandboxes), application hosting, serverless functions, scheduled jobs, and managed PostgreSQL databases, including storage, backup, and transmission of Customer Content.

Duration. The term of the agreement, plus the deletion window in Section 7.

Categories of data subjects and personal data. Determined by the customer. Customer Content may contain any category of personal data the customer chooses to process; PandaStack does not require or inspect it.

Purpose.Solely to provide, secure, and support the Services in accordance with the customer’s documented instructions.

PandaStack will:

• process Customer Content only on the customer’s documented instructions (the agreement, this DPA, and use of the Services’ APIs and settings constitute such instructions), unless required by law — in which case we will notify the customer unless legally prohibited;
• ensure persons authorized to process Customer Content are bound by confidentiality obligations;
• implement and maintain the technical and organizational measures in Annex 1, taking into account the state of the art and the risks of the processing;
• assist the customer, insofar as reasonably possible, in responding to data-subject requests (access, rectification, erasure, portability, restriction, objection) — the Services’ APIs let customers retrieve and delete Customer Content directly;
• assist the customer with security, breach-notification, impact-assessment, and consultation obligations under Articles 32–36 GDPR, taking into account the information available to us;
• make available information reasonably necessary to demonstrate compliance with Article 28, and allow for and contribute to audits as described in Section 8.

The customer grants general authorization for the subprocessors listed at pandastack.ai/subprocessors. PandaStack imposes data-protection obligations on each subprocessor materially no less protective than this DPA and remains liable for their performance. Customers who have executed this DPA may subscribe to advance notice of subprocessor changes and may object on reasonable data-protection grounds; if we cannot accommodate the objection, the customer may terminate the affected Services.

PandaStack will notify the customer without undue delay after becoming aware of a personal data breach affecting Customer Content, and will provide information reasonably available to us to help the customer meet its own notification obligations, followed by timely updates as the investigation progresses.

Customer Content is processed primarily in the United States. Where the GDPR or UK GDPR applies to a transfer, the parties incorporate by reference the European Commission’s Standard Contractual Clauses (Module 2: controller-to-processor, or Module 3 where the customer is itself a processor), and the UK International Data Transfer Addendum where applicable, with PandaStack as data importer and the customer as data exporter. Annex 1 serves as the description of technical and organizational measures for those clauses.

Customers can delete Customer Content at any time through the Services’ APIs and dashboard. Deleting a resource deletes its stored content: deleting a managed database also deletes its backup archive (base backups and WAL) from object storage. On termination of the agreement, PandaStack will delete remaining Customer Content within 30 days, except where retention is required by law.

PandaStack will make available the information reasonably necessary to demonstrate compliance with this DPA — including our security documentation, completed security questionnaires (CAIQ), and summaries of internal reviews. Where these are insufficient to meet a legal requirement, the customer may conduct (at its cost, on reasonable notice, no more than annually, and without access to other customers’ data) an audit confined to PandaStack’s processing of its Customer Content. The open-source core of the platform is additionally publicly auditable.

• Workload isolation. Every sandbox, app, function run, and managed database executes in its own Firecracker microVM with a dedicated Linux kernel on hardware virtualization (KVM); tenants never share a kernel. Each microVM runs as an unprivileged, seccomp-confined host process in its own network namespace.
• Encryption in transit. TLS for API, dashboard, app, and database traffic; managed database connections require TLS.
• Encryption at rest. Disks, snapshots, backups, and artifacts reside on cloud-provider storage with encryption at rest; application secrets are additionally encrypted with AES-256-GCM before storage and masked in reads and logs.
• Access control. Organization-scoped tenancy with role-based membership (owner/admin/member); API access via per-organization tokens stored as salted hashes; per-database credentials rotatable on demand; optional per-database IP allow lists and connection rate limits.
• Availability and recoverability. Managed databases receive automatic daily base backups plus continuous WAL archiving to durable object storage, with point-in-time recovery, restore, clone, and failover paths; archive integrity is verified by an automated scrubber whose result is exposed to the customer.
• Deletion. Resource deletion removes stored content, including database backup archives (see Section 7).
• Vulnerability management. A published security policy and disclosure process (security.txt, reporting guidelines); dependency and code review on an open-source core.
• Personnel. Access to production systems is restricted to authorized personnel bound by confidentiality, on least-privilege credentials.