Darcy Dispatch Data Processing Subprocessors
Last Updated
Version 10012026 — Effective October 1, 2026
This page (the “Subprocessor Page”) identifies the Subprocessors that Innovative Data Processing Solutions LLC (“Innovative”) engages to process Customer Personal Data in providing the MMR Services. It is maintained under Section 7.3 of the Darcy Dispatch Data Processing Addendum (the “DPA”) and forms part of the Agreement. Capitalized terms not defined on this page have the meanings given in the DPA or the Agreement.
1. SCOPE
1.1 What This Page Covers. A Subprocessor is any third party that Innovative engages to process Customer Personal Data, meaning Personal Data contained in the prompts Customer submits to Darcy Dispatch and in the outputs returned to Customer. Each Subprocessor listed on this page may receive Customer Personal Data in the course of serving a Metered Request.
1.2 What This Page Does Not Cover. Innovative processes Account Data and Service Records as an independent business under Section 2.2 of the DPA. Vendors that receive only Account Data or Service Records, such as providers of billing, metering, analytics, observability, and database services that store usage records but not the text of prompts or outputs, are not Subprocessors and are not listed on this page. Their processing is described in Innovative’s privacy policy.
1.3 Location. Every Subprocessor processes Customer Personal Data in the United States only.
2. PROVIDER STANDARDS
Each inference router and Underlying Model Provider listed on this page is engaged under written terms that meet the Provider Standards in Section 9.4 of the MMR Terms, including:
2.1 No Training. Customer Personal Data and Customer Content are not used to train, fine-tune, or otherwise improve any artificial intelligence or machine learning model.
2.2 U.S.-Only Inference and Storage. Customer Personal Data is processed and, where stored at all, stored in the United States.
2.3 Zero Data Retention. Prompts and outputs are not retained after the response is returned, other than transiently as needed to process the request or as required by law.
2.4 Restricted-Party Screening. No Subprocessor is based in, or controlled from, a country of concern under 28 C.F.R. Part 202.
3. INFERENCE ROUTERS
Inference routers receive prompts and outputs from Darcy Dispatch and transmit them to an Underlying Model Provider. They do not host model weights.
Subprocessor | Purpose | Customer Personal Data | Location |
|---|---|---|---|
OpenRouter, Inc. (OpenRouter) | Routing Metered Requests to Underlying Model Providers, enforcing Zero Data Retention routing. | Prompts and outputs, processed transiently. | United States |
Ramp Business Corporation (Ramp Router) | Routing Metered Requests to Underlying Model Providers, enforcing Zero Data Retention routing. | Prompts and outputs, processed transiently. | United States |
3.1 Additional Inference Routers. Innovative may engage additional inference routers that meet the Provider Standards and will add each one to this page before it processes Customer Personal Data.
4. UNDERLYING MODEL PROVIDERS
Underlying Model Providers perform AI model inference on prompts and return outputs. Prompts and outputs reach them through an inference router listed in Section 3, or directly where a router is not used.
Subprocessor | Purpose | Customer Personal Data | Location |
|---|---|---|---|
OpenAI, L.L.C. | AI model inference. | Prompts and outputs, processed transiently. | United States |
Anthropic, PBC | AI model inference. | Prompts and outputs, processed transiently. | United States |
Amazon Web Services, Inc. (Amazon Bedrock) | AI model inference. | Prompts and outputs, processed transiently. | United States |
Microsoft Corporation (Azure AI Foundry) | AI model inference. | Prompts and outputs, processed transiently. | United States |
4.1 Other Underlying Model Providers. Darcy Dispatch may route requests to other providers of AI model inference that meet the Provider Standards. Because Underlying Model Providers change frequently, providers not listed in the table above are identified by category only, as described in Section 7.3 of the DPA.
5. SUPPORTING SERVICES
Subprocessor | Purpose | Customer Personal Data | Location |
|---|---|---|---|
TypeSafe (Jev) | AI model inference supporting features of Darcy Dispatch. | An excerpt of the conversation in a request, and in some cases its system instructions, processed transiently. Classification results are recorded as Service Records and do not contain the text of prompts or outputs. | United States |
6. CLOUD INFRASTRUCTURE
Subprocessor | Purpose | Customer Personal Data | Location |
|---|---|---|---|
Amazon Web Services, Inc. | Hosting of Darcy Dispatch, including compute, key and configuration storage, message queuing, secrets and encryption key management, and security logging. | Prompts and outputs, processed transiently in memory while a Metered Request is served. Prompts and outputs are not written to persistent storage, except as described in Section 5.1 of the DPA. | United States |
7. CHANGES AND OBJECTIONS
7.1 Updates. Innovative may change its Subprocessors at any time and will update this page to reflect each change, including the Last Updated date above.
7.2 Objections. To the extent U.S. Privacy Laws require it, Customer may object to a change on reasonable data protection grounds by written notice within fifteen (15) days after this page is updated, as described in Section 7.4 of the DPA. The parties will discuss the objection in good faith. If they cannot resolve it, Customer’s sole remedy is to stop using, or terminate, the affected MMR Services, and no Early Termination Fee or Shortfall Charge applies to a termination under that Section.
7.3 Responsibility. Innovative remains responsible for its Subprocessors as described in Section 7.5 of the DPA.
