Data Processing Agreement
Version 1.2 · Effective 14 October 2026
Between the Customer (the entity that has registered for, or is accessing, the Navige service, acting as Data Controller) and Navige Ltd, a company registered in the United Kingdom (acting as Data Processor).
This Data Processing Agreement (“DPA”) forms part of the agreement governing the Customer’s use of the Navige service. By creating a Navige account or using the Navige API, the Customer agrees to this DPA. Where a separate written contract exists between the parties, this DPA supplements that contract and, in the event of conflict on data protection matters, this DPA prevails.
Article 1 — Definitions
In this DPA, the following terms have the meanings set out below:
Article 2 — Subject Matter, Duration, and Purpose
Article 3 — Obligations of the Controller
Article 4 — Obligations of the Processor
Article 5 — Sub-processors
Article 6 — Confidentiality
Article 7 — Personal Data Breaches
Article 8 — International Transfers
Article 9 — Data Subject Rights
Article 10 — Retention, Deletion and Return of Data
Article 11 — Audit Rights
Article 12 — Liability
Article 13 — General
Article 14 — Contact
Schedule 1 — Details of Processing
Types of Personal Data
- Names, job titles, and work email addresses of employees and contractors
- Account registration data: name, company, and work email of the account holder, and a hashed password where the account was created with one
- Where the account holder signs in with a third-party identity provider: the identifier, name and email address returned by that provider
- Redacted content of AI tool call arguments and parameters, which may contain personal data of employees or third parties. Masking is applied before storage, as described in Schedule 3
- Timestamps, AI platform identifiers, model names, agent identifiers and correlation identifiers
- Where provider billing is connected: the identity of the account owner who enabled it, and the credential they supplied, held encrypted. The cost figures themselves are organisational billing records rather than Personal Data, and contain no call content
Categories of Data Subjects
- Employees and contractors of the Controller who operate or administer AI agents and applications
- Third parties whose personal data may be referenced within AI tool call arguments, to the extent it survives masking
- Account holders and administrators who manage the Navige account
Data Navige does not process
- The text of prompts and completions passed through the model proxy
- Special category data, which the Services are not designed to process and must not be sent to them
- Payment card data, which is collected and held by Stripe and never reaches Navige’s systems
Schedule 2 — Approved Sub-processors
Current Sub-processors — last updated 14 October 2026.
| Sub-processor | Role | Location | Transfer mechanism |
|---|---|---|---|
| Railway Corp | Cloud infrastructure, compute, and database storage (PostgreSQL and Redis). Hosts the Navige gateway and console, and stores all audit log, governance, spend and account data. | United States | SCCs / UK IDTA |
| Resend Inc | Transactional email delivery: account verification, welcome and billing messages, weekly reports, and human-approval notifications. | United States | SCCs / UK IDTA |
| Stripe Inc | Payment processing, billing and subscription management. Stripe holds payment card data; it does not receive tool call data, audit log content, or spend records. | United States | SCCs / UK IDTA |
| Framer B.V. | Hosting of the public website at navige.ai, including the page that renders the sign-up form. Data entered into that form is submitted directly to Navige’s own systems and is not received or stored by Framer. Framer processes website visitor data, such as IP address and page requests, in its capacity as host. | Netherlands / EU | EU — no transfer safeguard required |
Identity providers. Where a user chooses to sign in with Google, Microsoft, GitHub or Slack, or through the Controller’s own SAML or OIDC provider, that provider acts as an independent controller of the sign-in it performs. Navige receives the identifier, name and email it returns. No Personal Data is processed by that provider on Navige’s behalf.
Model providers. Calls routed through the Navige proxy are forwarded to the provider the Controller has chosen, using the Controller’s own credential. That provider processes the call for the Controller, under the Controller’s own agreement with it, not on Navige’s behalf. The same applies to the optional provider billing retrieval under clause 2.5.
Polygon. For Controllers who enable blockchain anchoring, the Polygon public blockchain receives only a keccak256 hash of each batch of audit records, chained to the hash of the batch before it. No Personal Data, call content or identifiable information is transmitted to or stored on the Polygon network.
Schedule 3 — Technical and Organisational Security Measures
Measures in place as of 14 October 2026.
Version history
- 1.2 — 14 October 2026. Retention stated accurately (10.1–10.3); international transfers restated against the actual hosting arrangement (Article 8); SCC modules identified (8.3); Anthropic removed from Schedule 2 and Framer added; contact details added (Article 14); no-training commitment (4.7) and prompt-content position (2.7) made explicit; Schedule 3 expanded to cover secret redaction, append-only records, failure behaviour and environment separation.
- 1.1 — 6 October 2026. Added optional provider billing reconciliation (2.5–2.6).
Navige Ltd, registered in the United Kingdom · hello@navige.ai