Subprocessors
Last updated: 2026 Q2
This page lists the subprocessors L2h.ai, Inc. (“L2H”) uses for corporate operations and the marketing website at l2h.ai. Customer deployments of Orchestrator, Chat, and the Enterprise Agent suite (for ServiceNow and VS Code) run inside the customer’s own AWS, Azure, or on-prem environment; L2H does not act as a subprocessor for the runtime data plane of those deployments. The LLM providers section below identifies the third-party LLM services a customer may elect to integrate from inside their own environment.
1. Corporate & site subprocessors
The following subprocessors process personal data L2H collects for its own marketing, sales, support, and corporate operations.
| Subprocessor | Purpose | Region | Transfer mechanism |
|---|---|---|---|
| Vercel Inc. | Hosting of the public marketing site at l2h.ai (static export, CDN edge delivery). | Global edge (primarily US) | SCCs (where applicable) |
| HubSpot, Inc. | CRM, marketing email, contact-form processing, and meeting scheduling for L2H sales and marketing operations. | United States (na2) | SCCs |
| Google LLC (Google Workspace) | Corporate email, document collaboration, calendar. | United States | SCCs |
| Google LLC (Google Analytics 4) | Aggregate website analytics (only when consent is given). | United States / EU | SCCs |
| Microsoft Corporation (Clarity) | Optional usability research (heatmaps, session replay) when enabled and consent is given. | United States | SCCs |
| LinkedIn Corporation | LinkedIn Insight Tag for marketing campaign measurement (only when consent is given). | United States / Ireland | SCCs |
| GitHub, Inc. | Source-code management for L2H product code. | United States | SCCs |
| Amazon Web Services, Inc. (AWS) | Cloud infrastructure for L2H corporate workloads. Customer L2H deployments on AWS run in the customer's own AWS account; L2H is not a subprocessor of that runtime. | United States (multi-region; varies) | SCCs (where applicable) |
| Microsoft Corporation (Azure) | Optional cloud infrastructure for L2H corporate workloads. Customer L2H deployments on Azure run in the customer's own Azure subscription; L2H is not a subprocessor of that runtime. | United States / Customer GovCloud (varies) | SCCs (where applicable) |
2. LLM providers (customer-elected)
When a customer configures an L2H deployment to call a third-party LLM, that LLM provider becomes a sub-processor of the customer (not of L2H) for the model inputs and outputs the customer routes to it. Each provider operates under its own published terms, data-processing addendum, and applicable transfer mechanism. Customers should review the relevant provider’s documentation before enabling it. The list below is non-exhaustive.
| Subprocessor | Purpose | Region | Transfer mechanism |
|---|---|---|---|
| OpenAI OpCo, LLC | LLM provider, used in the customer environment when the customer configures it. | United States | SCCs |
| Anthropic, PBC | LLM provider, used in the customer environment when the customer configures it. | United States | SCCs |
| Google LLC (Google AI / Gemini) | LLM provider, used in the customer environment when the customer configures it. | United States / EU | SCCs |
| Amazon Web Services, Inc. (Amazon Bedrock) | Multi-model LLM gateway, used in the customer environment when the customer configures it. | Per customer-selected AWS region | SCCs (where applicable) |
| Microsoft Corporation (Azure OpenAI Service) | LLM provider, used in the customer environment when the customer configures it. GovCloud regions available. | Per customer-selected Azure region | SCCs (where applicable) |
| X.AI Corp. (xAI) | LLM provider, used in the customer environment when the customer configures it. | United States | SCCs |
| Self-hosted / open-source models (vLLM, llama.cpp, Ollama, OpenAI-compatible endpoints) | Run inside the customer's own infrastructure; no external transfer. | Customer-controlled | Not applicable |
3. Subscribe to updates
Customers can subscribe to subprocessor change notifications by emailing support@l2h.ai. We provide at least 30 days’ advance notice before adding a new corporate subprocessor that processes customer personal data, in accordance with our DPA.
4. Trademarks
Vendor names listed above are trademarks of their respective owners; reference does not imply endorsement, sponsorship, or affiliation. See the trademark notice.