Protecting your data is a core principle of how Helios operates. Whether you're uploading internal documents, asking Horizon about supplier strategies, or integrating our forecasts into your workflows, we never use your data in ways that compromise privacy, confidentiality, or control.
Here’s what you can expect:
No Selling or Sharing of Customer Data: We do not sell, share, or monetize customer data. Full stop. Your queries, uploads, and usage stay entirely within your organization’s account. No third parties. No internal cross-training. No marketing backchannels.
No Model Training on Your Inputs: We never use your data—documents, prompts, conversations, or API calls—to train our models. Our AI models are trained exclusively on trusted, pre-approved sources and proprietary datasets developed by Helios—not on anything you or your team say, do, or upload. Your proprietary information stays yours.
Enterprise-Grade Security Infrastructure: Helios is hosted on Google Cloud Platform, benefiting from best-in-class cloud security practices:
Encryption in transit and at rest
Role-Based Access Control (RBAC) and Principle of Least Privilege
Audit logging and real-time monitoring
Regular penetration testing and compliance reviews
Horizon is designed for sensitive internal questions
Helios and Horizon are secured end-to-end specifically so teams can ask confidential, high-stakes questions without worrying about data leakage. All traffic is protected with TLS (HTTPS); sensitive data is encrypted at rest; and every session is authenticated and scoped by role-based access so users only see what they’re entitled to. We also apply tenant isolation, rate-limiting to prevent abuse or bulk extraction, and—crucially—we do not train our models on your proprietary data (including anything you upload or type). Your conversations, uploads, and prompts stay your property and are used only to serve your workspace.
On the infrastructure side, Helios runs on Google Cloud Platform (GCP) for hardened security and compliance. For chatbot features that call Google’s Gemini API, your prompts and responses are not used to improve Google’s models and are processed under Google’s Data Processing Addendum; logs related to safety/compliance may be retained briefly to detect policy violations, and operational metadata remains governed by Google’s standard privacy terms. We maintain continuous event logging and monitoring to detect suspicious activity, and we perform regular reviews and patching to keep defenses current. Bottom line: from transport to storage to third-party integrations, every layer is designed for confidentiality, integrity, and resilience, so you can safely use Helios for sensitive internal analysis, strategy, and decision support.