Terms of Service Privacy Policy Acceptable Use Policy Subprocessors

Newtown Platform — Data Processing Addendum

Effective date: 2026-10-13 Version: 1.3

This Data Processing Addendum ("DPA") forms part of the Newtown Platform Terms of Service (the "Agreement") between Newtown Analytics LLC ("Newtown," "Processor") and the customer that accepted the Agreement ("Customer," "Controller"). Where the Agreement and this DPA conflict on the processing of Personal Data, this DPA governs — including over an Order Form.

This DPA takes effect with the Agreement and requires no separate signature. If your procurement process needs a countersigned copy, email legal@newtowndata.io.

1. Definitions

• "Data Protection Laws" means all laws applicable to the processing of Personal Data under this DPA, including the EU General Data Protection Regulation 2016/679 ("GDPR"), the UK GDPR and Data Protection Act 2018, the California Consumer Privacy Act as amended ("CCPA"), and comparable U.S. state privacy laws.
• "Personal Data" means personal data, personal information, or personally identifiable information within Customer Data.
• "Customer Data," "Service," "Data Source," "Warehouse," "AI Client," "Workspace," "Order Form," and "Output" have the meanings given in the Agreement.
• "Sub-processor" means a third party engaged by Newtown to process Personal Data.
• "Security Incident" means a breach of security leading to the accidental or unlawful destruction, loss, alteration, or unauthorized disclosure of or access to Personal Data processed by Newtown.
• "Controller," "Processor," "Data Subject," "Processing," and "Supervisory Authority" have the meanings given in the GDPR. Under the CCPA, Customer is the "business" and Newtown is a "service provider."
• "Standard Contractual Clauses" or "SCCs" means the clauses annexed to European Commission Implementing Decision (EU) 2021/914, and where relevant the UK International Data Transfer Addendum ("UK Addendum").

2. Roles and scope

Customer is the Controller of Personal Data within Customer Data. Newtown is the Processor. Each party complies with the Data Protection Laws applicable to it in that role.

Newtown is a Controller in respect of the account data described in the Privacy Policy (workspace email addresses, invite records, billing contacts, correspondence, and operational telemetry). This DPA does not govern that data; the Privacy Policy does, including its commitment to notify Customer of a breach affecting it.

Customer's responsibilities. Customer determines which Data Sources to connect and for what purposes, and is solely responsible for: the lawfulness of the collection and of the transfer to Newtown; providing notices and obtaining consents required from Data Subjects; the accuracy of Customer Data; and the legality of Customer's instructions. Customer warrants that connecting a Data Source breaches no agreement it has with that source or with any Data Subject.

3. Processing instructions

Newtown processes Personal Data only:

1. on Customer's documented instructions, comprising the Agreement, this DPA, and Customer's configuration and use of the Service — including which sources it connects, which models it requests, and which AI Client it connects;
2. as necessary to provide, secure, operate, and support the Service and to respond to support requests; and
3. where required by law, in which case Newtown will inform Customer before processing unless the law forbids it; and
4. where Customer has selected Newtown-Managed Transformation in a warehouse Customer owns, by creating, altering and deleting datasets, schemas, tables and views in that warehouse as necessary to build, maintain, validate and document Customer's data model, on a recurring schedule, under a write identity separate from the read identity and within the grants Customer has made. Newtown does not alter Customer's source data.

Newtown will tell Customer if, in its opinion, an instruction infringes Data Protection Laws.

Restrictions Newtown accepts. Newtown will not: sell or share Personal Data (as defined in the CCPA); retain, use, or disclose Personal Data outside the direct business relationship with Customer or for any purpose other than performing the Service; combine Personal Data with data from another source except as the CCPA permits a service provider; or use Personal Data to train, fine-tune, or develop machine-learning models. Newtown's agreement with its AI Sub-processor prohibits that Sub-processor from training on data Newtown submits.

Newtown may create and use aggregated and de-identified data derived from the processing, provided it cannot reasonably be used to identify Customer or any Data Subject, and will not attempt to re-identify it.

4. Confidentiality and personnel

Newtown limits access to Personal Data to personnel who need it to perform the Service. All personnel and contractors with access are bound by written confidentiality obligations that survive their engagement.

Staff access is controlled by Customer. For Workspaces created after staff access controls were introduced, Newtown staff access to Workspace content is off by default; Customer can turn it on, off, or on for 1 hour, 24 hours, or 7 days from the console. Workspaces that existed before then have access on and Customer may switch it off. While enabled, every staff access to Workspace content is logged and Customer can view that log in the console.

Exceptions that apply regardless of the setting: (a) automated platform operations — nightly dbt builds, nightly modeling and surveying agent runs, connector health checks, dormancy and billing notices; (b) operational, billing, and provisioning metadata containing no Workspace content (tenant name, plan, status, usage, health); and (c) access reasonably necessary to protect the security or integrity of the Service or to comply with law.

Customer should be aware: infrastructure-level administrators of Newtown's own cloud environment can technically reach stored data, including Customer's. The Customer setting and log do not cover that access; it is governed by written confidentiality obligations, Newtown's internal access controls, and cloud audit logs, which are not provided to Customer.

5. Sub-processors

General authorization. Customer authorizes Newtown to engage Sub-processors. The current list is at [subprocessors.md](./subprocessors.md), incorporated into this DPA.

Flow-down and responsibility. Newtown enters into a written agreement with each Sub-processor imposing data protection obligations no less protective than this DPA, and is responsible for each Sub-processor's processing of Personal Data to the same extent as for its own, subject to the *Limitation of Liability* section of the Agreement.

Changes. Newtown will give at least 30 days' notice before adding or replacing a Sub-processor, to Customer's Notice Address and by updating the subprocessor list. Customer may object on reasonable data-protection grounds within that period. The parties will work in good faith to resolve the objection; if they cannot, Customer may terminate the affected subscription without penalty, effective at the end of the then-current billing period, with a pro-rata refund of prepaid, unused fees. That is Customer's sole remedy for an objection. Newtown will maintain a notification list for subprocessor changes; keeping your subscription to it current is your responsibility.

Not a Sub-processor: Customer's AI Client. Claude, ChatGPT, or another product Customer connects to the MCP endpoint is not a Newtown Sub-processor. Newtown does not select it, does not contract with it, and cannot impose terms on it. When Customer connects one, query results leave the Service to that provider under Customer's own account. Customer is the Controller for that transfer, is responsible for its lawfulness, and is responsible for putting Article 28 terms in place with that provider where it acts as Customer's processor.

6. Security

Newtown implements the technical and organizational measures in Annex II, taking into account the state of the art, the cost of implementation, and the nature, scope, and risks of the processing.

Annex II describes the Service as it actually exists. Newtown holds no SOC 2, ISO 27001, or PCI certification and will not represent otherwise. Newtown may update Annex II provided the overall level of security is not materially reduced. Additional detail about Newtown's security posture, including items not yet implemented, is available in Newtown's security overview on request; that document is informational and not contractual.

7. Security Incidents

Newtown will notify Customer without undue delay, and in any event within 72 hours of confirming a Security Incident affecting Personal Data processed on Customer's behalf.

The notification will describe, to the extent known: the nature of the incident, the categories and approximate volume of Personal Data and Data Subjects affected, the likely consequences, and the measures taken or proposed. Where that is not all available at once, Newtown will provide it in phases without undue further delay.

Newtown will take reasonable steps to contain and remediate the incident and will cooperate with Customer's investigation and with any notifications Customer must make to a Supervisory Authority or to Data Subjects. Notification is not an acknowledgment of fault or liability.

Send security reports to security@newtowndata.io.

8. Assistance to Customer

Taking into account the nature of the processing and the information available to it, Newtown will assist Customer:

• Data Subject requests. The Service gives Customer direct access to its own data through the Warehouse, the dbt project, and export, and Customer should use that access to satisfy requests. If a Data Subject contacts Newtown directly, Newtown will not respond substantively, will refer them to Customer, and will notify Customer promptly. Where Customer cannot satisfy a request through the Service, Newtown will provide reasonable assistance.
• DPIAs and prior consultation. Newtown will provide reasonably available information to support Customer's data protection impact assessments and any prior consultation with a Supervisory Authority.
• Security questionnaires. Newtown will respond to reasonable security and privacy questionnaires, no more than once per year absent a Security Incident or a material change.

Assistance that is disproportionate or repetitive is billed at $250 per hour, quoted and agreed in advance.

9. Audit

Newtown will make available the information reasonably necessary to demonstrate compliance with this DPA, and will allow and contribute to audits, including inspections, conducted by Customer or an auditor it mandates.

In practice, in escalating order: (1) Newtown responds with written documentation, Sub-processor certifications, and a completed questionnaire; (2) if that is genuinely insufficient, a remote audit on at least 30 days' notice; (3) an on-site inspection only where a Supervisory Authority requires it or following a Security Incident affecting Customer. Audits are limited to once in any 12 months unless a Supervisory Authority requires otherwise or a Security Incident has occurred, take place during business hours, must not unreasonably disrupt Newtown's operations, and are subject to confidentiality. Customer bears its own audit costs and reimburses Newtown's reasonable costs beyond the annual audit. Auditors must not be competitors of Newtown.

Audits must not extend to other customers' data or environments, or to Sub-processors' facilities beyond what those Sub-processors permit.

10. Deletion and return

On termination or expiry, Newtown handles Personal Data in accordance with the *Export, Deletion, and Continuity* section of the Agreement: a 30-day post-termination export window (60 days under a negotiated Order Form, and no export window for a trial that did not convert), after which Newtown suspends the Workspace. Newtown deletes three months after termination or expiry, and earlier on Customer's written request. Deletion covers the Workspace and, for a Newtown-provisioned Warehouse, the underlying cloud project, together with the ingestion connections and destination, the generated repository, the business context and compiled documentation, and the Workspace's core control-plane records. Deletion is performed by Newtown staff rather than by an automated job. Closing a Workspace suspends access and disables credentials; Newtown can restore it on Customer's request for 30 days after closure, and permanent deletion (secrets, service accounts, connections, repository, cloud project) is performed no earlier than 30 days after closure unless Customer requests earlier deletion in writing. The three-month period above remains the outer date; the 30-day restore period is a minimum and not a backup or recovery commitment.

What is not deleted automatically, and which Newtown will delete on written request: invite records, including the email address of the person who redeemed the invite; authorization records associated with machine access to the Workspace; and any source credential Customer supplied to Newtown directly. Newtown may also retain Personal Data where required by law, and copies persist in backups and system logs until they age out on their normal schedule. Anything retained remains subject to this DPA and will not be processed for any purpose other than completing the deletion or meeting the legal requirement.

Where Customer uses a Bring-Your-Own Warehouse, Newtown holds no copy of the underlying data; on termination it revokes access, deletes the credentials it held, and deletes its control-plane records. Where Newtown has maintained a data model inside that warehouse under Newtown-Managed Transformation, the Target Schemas Newtown created and their contents remain in Customer's warehouse under Customer's control — Newtown ceases writing to them and does not delete them — and the repository holding the generated project is transferred or delivered to Customer in accordance with the *Export, Deletion, and Continuity* section of the Agreement.

Newtown will confirm in writing what has been deleted, on request.

11. International transfers

Newtown processes Personal Data in the United States.

Where Customer transfers Personal Data subject to the GDPR or UK GDPR to Newtown and no adequacy decision applies:

• The SCCs are incorporated by reference, with Module Two (Controller to Processor) where Customer is a controller and Module Three (Processor to Processor) where Customer is itself a processor.
• Clause 7 (docking) applies. Clause 9 uses Option 2, general written authorization, with the 30-day notice period in Section 5. Clause 11's optional independent-dispute-resolution language does not apply. Clause 17 selects the law of Ireland; Clause 18(b) selects the courts of Ireland.
• Annex I to the SCCs is populated by Annex I below; Annex II to the SCCs by Annex II below; the Sub-processor list by the subprocessor annex.
• For UK transfers the UK Addendum applies, with Tables 1–3 populated by this DPA and its Annexes and Table 4 selecting "neither party" under Section 19.
• For Swiss transfers, references to the GDPR are read as references to the Swiss FADP, references to Member State law as Swiss law, and the Swiss FDPIC is a competent Supervisory Authority.

Newtown will notify Customer if it becomes unable to comply with the SCCs, and will challenge any government request for Personal Data it believes unlawful.

12. CCPA and U.S. state laws

Newtown acts as a service provider (CCPA) / processor (other U.S. state laws) and certifies that it understands and will comply with the restrictions in Section 3. Newtown does not sell or share Personal Data and receives no consideration for it beyond the fees for the Service.

13. Liability

Each party's liability under this DPA is subject to the limitations and exclusions in the *Limitation of Liability* section of the Agreement, including the enhanced cap that applies to a breach of this DPA or a Security Incident caused by Newtown. Nothing in this DPA limits a Data Subject's rights under Data Protection Laws or a party's liability to a Supervisory Authority.

14. General

This DPA is governed by the law stated in the Agreement, except that the SCCs are governed as stated in Section 11. If a provision is invalid, the rest survives. This DPA supersedes any prior data processing terms between the parties for the Service.

Annex I — Details of the Processing

A. List of parties

• Data exporter / Controller: the Customer that accepted the Agreement. Contact: Customer's Notice Address. Activities: use of the Newtown Platform to ingest, model, and query its business data. Role: Controller, or Processor where Customer processes on behalf of its own clients.
• Data importer / Processor: Newtown Analytics LLC, New York, United States. Contact: privacy@newtowndata.io. Activities: provision of the Newtown Platform. Role: Processor.

B. Description of the transfer

*Categories of Data Subjects.* Determined by Customer through its choice of Data Sources. Typically: Customer's customers and prospects; Customer's employees, contractors, and users; Customer's suppliers and their personnel; and Customer's Workspace users.

*Categories of Personal Data.* Determined by Customer. Typically: name; business and personal email address; telephone number; billing and shipping address; company and job title; customer, subscriber, and account identifiers; transaction, invoice, charge, refund, and subscription records; order and product data; support and communication metadata; and technical identifiers present in source tables. Where Customer connects an accounting, payroll, banking, or payment connector, also: bank account and routing numbers, payroll and compensation records, tax identifiers, government-issued identification numbers, and cardholder data as exposed by that connector. Customer's Workspace users: work email address and email domain.

*Sensitive data.* The Agreement and the Acceptable Use Policy prohibit Customer from connecting sources carrying protected health information (absent a Business Associate Agreement), data about children under 13 (or under 16 where local law requires consent), export-controlled or classified or controlled government material, and any data Customer does not have the right to have processed. Government identification numbers, biometric identifiers, precise geolocation, financial account numbers, and cardholder data are permitted where they arrive through a standard source connector, and may therefore be present.

These are contractual restrictions only. Newtown applies no technical control that detects, blocks, or specially safeguards any category of data, and the connector catalogue offered is not restricted by category. Customer is responsible for assessing the risk of each source it connects and for notifying Newtown before connecting a payroll or bank-feed source so that column-level exclusions can be agreed.

*Frequency.* Continuous for the duration of the Agreement — scheduled ingestion syncs, scheduled transformation builds, and on-demand queries.

*Nature and purpose.* Ingestion from Customer's Data Sources into a Warehouse; storage; transformation into staging and modeled datasets by dbt; generation of metric definitions, documentation, and business context, including, only for tables where Customer has enabled "Sample values when profiling" (off by default), sampling of values from low-cardinality columns and retention of example join-key values for relationship verification; execution of queries against the datasets and return of results to Customer's AI Client; and provision of support.

Where Customer owns the Warehouse and has selected Newtown-Managed Transformation: creation, alteration and deletion of datasets, schemas, tables and views within Customer's own warehouse as necessary to build, maintain, validate and document Customer's data model, and recurring transformation of Customer's source data into them, under a dedicated write identity separate from the read identity and within the grants Customer has made. Newtown does not alter Customer's source data.

*Duration.* The term of the Agreement, plus the post-termination export window, plus the residual periods described in Section 10.

*Sub-processors.* See [subprocessors.md](./subprocessors.md), each for the duration of the Agreement.

C. Competent Supervisory Authority. Determined under Clause 13 of the SCCs by reference to the exporter's establishment or its Article 27 representative.

Annex II — Technical and Organizational Measures

These measures describe the Service as built. Newtown does not claim a measure it does not operate.

Encryption

• In transit: every production endpoint is served over HTTPS by the cloud provider's managed front end. The application does not itself enforce a redirect from plaintext HTTP; transport security is a property of the managed platform on which it runs.
• At rest: all storage — warehouse datasets, control-plane database, object storage, and secrets — is encrypted by default using the cloud provider's platform encryption with provider-managed keys. Customer-managed encryption keys (CMEK) are not offered.

Access control and identity

• Customer sign-in is federated to Google and requires a verified work-email address, screened against a non-exhaustive list of known consumer and disposable domains; the invite requirement rather than that list is the gate. Newtown never handles or stores a customer password, and requests only the openid email scope.
• Workspace membership is verified server-side from Google's signed token and never trusted from the browser. Sessions are signed cookies expiring after 24 hours. The session cookie is signed, not encrypted.
• Machine access to the MCP endpoint is available over OAuth 2.1 with PKCE, dynamic client registration, refresh-token rotation, revocation, and resource-scoped tokens; the user approves a consent screen naming the client, sessions are re-validated at refresh and end after 90 days, and removing a person from a Workspace allow-list ends their connector access within an hour. A per-Workspace URL path token is also accepted; it is bearer-equivalent, appears in server request logs by virtue of being in the URL, and is rotated on request. Newtown recommends OAuth.
• Internal administrative functions are restricted to Newtown staff email accounts. Where the deployment is configured with separate application and administrative hostnames, the two are separated at the hostname layer; where a single hostname is configured, separation is enforced at the authorization layer only.
• Token comparison is constant-time and unknown Workspaces and invalid tokens return identical responses, so a Workspace cannot be enumerated.
• Not offered: per-user roles or permissions within a Workspace; per-user provisioning — membership is by verified email domain once a Workspace is associated with one, and an allow-list of named addresses is available on written request; SSO to a non-Google identity provider; a customer-visible log of anything other than Newtown staff access to Workspace content. Multi-factor authentication depends on Customer's own Google Workspace configuration, which Customer controls.

Tenant isolation

• Each customer is provisioned a dedicated cloud project containing its own warehouse datasets. There is no shared multi-tenant dataset, and a query authorized for one Workspace cannot reach another Workspace's data.
• The identities holding roles on a tenant project are: the ingestion loader service account, scoped to dataset-editor and job-user on that project only; Newtown's provisioning service account, which holds project-level administrative roles because it creates the datasets, the loader account, and the quota; and Newtown's agent service account, scoped to read. Newtown staff hold no standing individual grants on a tenant project; administrative roles are held at the service-account and folder level, and a staff member self-grants a specific role at the folder level when performing a documented operation such as moving or deleting a project.
• Ingestion is isolated per customer: a distinct connector group and destination per Workspace.
• Application-layer authorization restricts a signed-in customer to paths naming its own Workspace.
• Where Newtown writes into a warehouse the Customer owns, it does so under an identity distinct from the one it uses to read, and the separation is enforced by the warehouse's own grant model rather than by Newtown's naming conventions. The write identity's authority to create and maintain schemas does not carry read access to data the Customer has not separately granted Newtown to read.
• Shared surfaces, disclosed: warehouse isolation is per project; control-plane isolation is at the application layer. Newtown operates one ingestion-provider account, one code-hosting organization and app installation, one control-plane database instance, one secret store, and shared object-storage buckets for business context and compiled documentation separated by per-Workspace object prefix.

Secrets management

• Warehouse and source credentials are stored only in the cloud provider's secret manager. By design they are never written to the control-plane database, never emitted to a log, and never returned in an API response or read back to a client.
• Source credentials for connectors are entered by Customer directly into the ingestion provider's own authorization page and never pass through Newtown's infrastructure, with one exception: for sources Newtown supports directly (currently Stripe), Customer supplies a restricted-scope key, which Newtown checks for the read access the loader needs and stores in the secret manager. Full secret keys are rejected; Newtown does not independently enumerate a restricted key's write permissions, and issuing the key with read-only scopes is Customer's responsibility.
• The service-account key used as the ingestion provider's warehouse loader credential is created by Newtown and passed to that provider in memory; Newtown does not persist it.
• This paragraph concerns credentials. The per-Workspace URL path token is handled as described under *Access control*.

Data minimization in operations

• The staff warehouse inventory panel reads metadata only — dataset and table names, row counts, sizes, column counts — and does not display row contents.
• Two components read row values, only for tables where Customer has enabled "Sample values when profiling" (off by default for all Workspaces): the schema profiler samples values from low-cardinality text columns to generate the onboarding interview and business context, and relationship verification retains a small number of example join-key values. Where enabled, both are stored in the Workspace's context and may appear in prompts sent to the AI Sub-processor.
• Prompts sent to the AI Sub-processor, including on nightly scheduled modeling and surveying runs for every active Workspace, carry business context, schema, table and column names, metric definitions, model source, change-request text, and the SQL text of queries asked through Customer's AI Client (which can contain literal values). Sampled column values and example key values are included only where sampling is enabled. Computed metric values are not sent. Bulk row-level extracts are not included.

Query and cost controls

• Per query: a dry run estimates bytes scanned and the query is refused before execution if it exceeds the configured byte ceiling; the same ceiling is applied as a hard maximum-bytes-billed limit at execution. Results are capped at a configured row count, above which they are truncated and flagged as truncated. Execution is subject to a timeout.
• Per project: a daily bytes-billed quota is configured on each provisioned tenant project as a cost ceiling. It is applied on a best-effort basis at provisioning time and Newtown has not verified that it refuses a query at the limit.
• Per Workspace: agent processing is capped by a daily token limit, enforced before each call against a persisted counter. The counter is not lock-protected, so two concurrent runs against the same Workspace can undercount each other and the daily cap can be exceeded.
• These are cost controls. Nothing in the Service limits the number of queries a Customer or its AI Client may run.

Change management

• Automated agents produce data-model changes as pull requests, which are merged by a person or automatically as described in the Agreement. Automated merging is a configuration setting rather than a committed control, and where it is in use it applies only to a class of change that has already graduated — a configured number of consecutive clean outcomes for that class, reset by any single failure — and never to the changes that earned the graduation. Newtown does not commit to human review of any individual change. Review is an operating practice, not a technically enforced control: branch protection is not enabled on Workspace repositories, and continuous-integration lint is not a required status check.
• Generated SQL is linted in continuous integration against an executable style standard.
• An automated test suite runs before deployment.

Infrastructure and resilience

• Managed, patched platform services from a single major cloud provider. No customer-facing servers are hand-patched by Newtown.
• The application runs as a stateless autoscaling managed service; the control-plane database and object storage are managed services.
• Not in place, and material to Customer's own planning: a documented and restore-tested backup program; a published recovery point or recovery time objective; automated uptime monitoring and alerting; a public status page. Newtown makes no backup or recovery commitment. Warehouse contents are re-ingestible from Customer's sources and the generated project is held in version control; the accumulated business context is the one artifact not reproducible from elsewhere.

Organizational

• Written confidentiality obligations for all personnel and contractors with access, surviving their engagement.
• An internal policy restricting the handling of customer data in third-party AI tools.
• Access limited to personnel who need it for provisioning, support, and model review; staff access to Workspace content is off by default and Customer-controlled and logged as described in Section 4. Infrastructure-level administrator access to Newtown's cloud environment is outside that control.
• No SOC 2, ISO 27001, PCI DSS, or HIPAA certification or attestation.

Deletion

• Termination triggers suspension; deletion follows three months after termination, or earlier on Customer's written request. Deletion covers the Workspace, the ingestion connections and destination, the generated repository, the business context and compiled documentation, the Workspace's core control-plane records, and, for a Newtown-provisioned Warehouse, the cloud project holding the datasets. It is performed by Newtown staff rather than by an automated job.
• Closure suspends access and disables credentials; restoration is possible on request for 30 days after closure; permanent deletion is no earlier than 30 days after closure unless Customer requests earlier deletion in writing.
• Provider-level deletion of a cloud project completes after an approximately 30-day recoverable window.
• Further items deleted on request rather than automatically are listed in Section 10.

Contact: privacy@newtowndata.io · security@newtowndata.io · legal@newtowndata.io Newtown Analytics LLC · New York